TECHN ICAL BULLETIN NO. 88 FEBRUARY 1974

Soil Toxicities as Causes ofSugarcane Leaf Freckle,

Macadamia Leaf Chlorosis (Keaau),and Maui Sugarcane Growth Failure

HARRY F. CLEMENTS, E. W. PUTMAN,R. H. SUEHISA, G. L. N. VEE,

AND M. L. WEHLING

HAWAII AGRICULTURAL EXPERIMENT STATION, UNIVERSITY OF HAWAII

ACKNOWLEDGMENTS

The authors wish to thank the managements of Kilauea Sugar Compan y, th eRoyal Hawaiian Macadamia Nut Company at Keaau, and Hawaiian Commercial andSugar Company at Puunene for their generous coo pera tio n in providing soil s for thestudies. Thanks also are due Takashi Nonaka and his st aff at the Brewer Crop Lo gLaboratory for so me of the an alytical data.

Especial thanks go to Walt er S . P. Vee , He ad o f the Computing Center at theUniversity of Hawaii , and Bonnie Ching for th e many sta t ist ica l ana ly ses .

Fin all y, ver y sincere thanks go to the exe cutives and directors o f C. Brew er andCompan y, Ltd. , of Honolulu, for con tin uing to provide funds without wh ich suc hstudies as these co u ld not have been made.

University of HawaiiCollege of Tropical Agricul ture

Hawaii Agricultural Experiment StationTechnical Bulletin 88

ERRATUM

SOIL TOXICITIES AS CAUSES OF SUGARCANE LEAF FRECKLE,MACADAMIA LEAF CHLOROSIS (KEAAU),

AND MAUl SUGARCANE GROWTH FAILURE

Page 7, line 1:

Page 18, line 15:

line 24:

Page 31, line 4:

Page 37:

Page 40, line 13:

line 32:

line 34:

Read per pound of ammonium nitrogen instead ofper pound, ammonium nitrogen.

Read buffer curve instead of buffer curse.

Read adsorbed instead of absorbed.

Read V-Ca(N03h ,CaP; instead ofV-Ca(N03h mCaP.

Add superscript 2 after center head MAUlSUGARCANE GROWfH FAILURE".

Read bucket instead of basket.

Delete parentheses around SA and DAP.

Delete parentheses around CaP.

CONTENTS

2 125252628293 133363 73 73940

Pag e334477

11121216161820

REFERENCES CITED

INTRODUCTION .STATEMENT OF OBJECTIVES .COMPOSIT ION OF VARIOUS LAVAS .HISTORY .SUGARCANE LEAF FRECKLE .

Exp erimental Background .Usefu l Si licates and Lo ca l Sili cate Materials .The Caus es of Freckling in Sugarcane .

Analysi s of Ca ne Leaves .An alysis of Roots .Soil So lution Toxicity .Scanning and Resin Ex ch ange Columns .Specifi city of the Silica te Action .Effects o f Calcium Silicat e vs. Sili cic Acid and ofFerti lizers on Soil Toxicit y .

Id ent ify ing Areas in Need of Calcium Metasi licate .MACADAMIA LEAF CHLOROS IS AT KEAAU .

Foliar Analysi s Compari sons .Soil Studies .Total Analysis of Aqueo us Extract .So il Cu lture Stud ies .Effect of Treatments on Soil pH .Gen er al Conclusions and Recommendations .

MAUl SUGARCANE GROWTH FA ILURE .Tissue Analysis .So il Extracts .Le aching .Effec t of Treatments on th e Ab sorption o f th e Various Nutrientsby Sudangrass 45Soil pH Changes 48Ge neral Conclusions and R ecom m endat ion s 50

5 1

Figures

l ao Sugarcane p lants showing th e reddish bronz ing of o ld leaves as we llas th e freckling of young ones (H 53-263) 8

l b. V arious stages of lea f fr eck ling with two no rma l leaves ins ertedfor co n trast (H51 -2279) 8

2. Stages in the intensity o f fr eck ling 9

THE AUTHORS

BARR Y F. CLEMENTS is Senior Plant Physiologist, Emeritus, Hawaii Agricultural

Experiment Stat ion, and Consu lting Plant Physio log ist, Retired, C . Brewer an d Co .,Ltd. EDISON W. PUTMAN is Asso ciate Plant Physiologist, Haw aii Ag riculturalExperiment Station, and Asso ciate Pro fesso r of Plant Physiology, University ofHawaii. ROBERT H. SUEHISA is Junior Plant Physiologist, Hawaii AgriculturalExperiment Station . GLADYS L. N. VEE was fo rmer Assistant in Plant Physiology ,lIawaii Agricultural Experiment Sta t ion. MARGARET L. WEHLING was formerJunior Plant Physiologist, Hawaii Agricultural Experiment Station.

Soil Toxicities as Causes ofSugarcane Leaf Freckle,

Macadamia leaf Chlorosis (Keaau),and Maui Sugarcane Growth Failure

HARR Y F. CLEME NTS, E. W. PUTM A N,R . H. SU EHISA , G . L . N. YEE ,

and M. L. WEHLING

INTRODUCTION

Y ields of sugarcane receiving ade quate mineral nu triuon may be predicted byequations involving on ly sunligh t, maximum temperature, mi nim u m temperature,age o f the ca ne, an d ti ssu e mo isture levels for the period of calc ula t ion (7 ). Ex ce p tfo r the nutrients and water ad de d to them, intensively cropped soils, other thanserving as reservoirs, con tr ibu te little of a posit ive nature to crop y ield and of te nh ave negative effec ts . The most id eal soil wo u ld b e one wh ich detracts in no wayfro m the y ield potential of the particular site . Suc h a soil probably does no t ex ist.Many biotic factors in the so il probab ly are d ominant ca uses of reduced yields aswell as of permanent varietal de cline , even though othe r biotic fac tors are helpful.Lack of ade quate soil o xygen is probably as imp or tant a factor as lack of anynutrient or water. Solutes which exist in t he so il solu tion may be helpful as well asharmful. In very aci d clay so ils , many essential minor elemen ts are as fr equentlytoxi c as they are helpful.

So me non essenti al eleme n ts resulting fro m natural weat he ring can be very toxic.Oft en qu anti ti es o f suc h cleme n ts are increased in the so lutio n by th e fe rt ili za tionpract ices. Alt ho ug h this bullet in will co nc ern it sel f o nly with th e in organicto xicities, man y o f the organic so lu tes , in cluding the enti re array o f degradationproduct s res u lt ing from th e various decay cycles of biological re m na n ts , are to xic ,although so me are st imulative (31 , 32, 33, 34).

STATEMENT OF OBJECTIVESThe natural co mposi t io ns of so me of Haw aii 's crop soi ls d er ived from lavas are

ex plo re d to det ermine th e to xi cit ies ca use d by inorgan ic so lu tes, th e ir o r igin an dnature, th e ir effec ts on certain cro p s, an d methods of correc t io n . Part icularat ten ti on is di rect ed to three very seve re m aladies whi ch oc cu r in so ils co m mo n lyfo u nd in Haw aii : leaf freckle o f suga rca ne , macad am ia leaf ch lo rosis, an d Mauisuga rcane gro wt h failure.

4 HAWAII AGRICULTURAL EXPERIMENT STATION

COMPOSITION OF VARIOUS LAVASSince most of the soils in Hawaii are derived dire ctly or indirectly from volcanic

lavas, their co mp osit ion is coge nt at this point. Ma cdonald (12) reports the majorelemental co mp osit ion of several lavas (T able 1). There is an overall similarity,alt ho ugh variations among them undoubtedly help to explain differences cropscon tend with in various fields. Silica in the form o f silicat es is the dominantco nst itu en t. Other elements are alu mi nu m (as Al 20 3) , ranging from 9.1 to 19.8percent, ferric iron (as Fe2°3), from 2.6 to 9 .1 percent, and ferrou s iron (as FeO ),from 1.0 to lOA percent. The ordinary bases, Ca, Mg, K, an d Na as oxides, m ak e upanyw he re from 5.9 to 36.7 percent o f the sever al lavas. Man ganese oxide , ran gingfrom 0.1 3 to 0.24 percent, is ver y low cons idering it s impo rt an ce in so me aci d so ils.Titanium o xide makes up 1.1 to 4. 3 per cent , an d phosphorus pentox ide (P2 O s )from 0.2 to 204 percent.

Table I. Percentage composition of several lavas (aft er Macdonald [12 J )

Volcanic areas

West Maui Mauna KeaOxides Waianae and Molokai and Kohala Hualalai Koloa

Si02 45.9 -49.8 45.2 -58.0 45 .5 -51.8 45.6 -46.8 40.0 -45.4AI203 13.5 -17.3 13.3 -19.8 14.0 -17.1 13.9 -15 .3 9.1 -1 6.4Fe203 4.2 - 9.1 2.6 - 7.0 3.1 - 6.4 3. 1 - 4.6 3.0 - 7.0FeO 4.1 - 9.7 1.0 -10.4 6.3 -10. 1 8.1 -10.2 6.2 - 9.7MgO 3.6 - 5.3 1.0 - 6.5 2.4 -16.3 7.1 - 10.0 4.3 -15.9Cao 6.4 - 9.4 1.7 -10.3 5.0 -1l .8 9.6 -11.5 8.6 - 14.3Na20 3.0 - 4.6 2.8 - 6.2 1.5 - 5.8 2.4 - 3.2 2.1 - 5.3K20 I.l - 2.1 .4 - 3.0 .3 - 2.8 .9 - I.l .5 - 2.0H2O+ .33- .9 2 .2 - 2.1 .2 - .8 .2 - .6 .2 - 1.0H20 - .54- .96 .4 - 2.4 .2 - .7 .1 - .2 .4 - I.lTi02 3.0 - 4.3 l.l - 3.7 1.4 - 4.0 2.1 - 2.6 2.0 - 3.2P20S .2 - I.l .6 - 2.4 .3 - 2.0 .2 - .4 .4 - I.7MnO .13 - .18 .15- .24 .16- .22 .16- .18 .15- .19

HISTORYMcG eorge (19), as early as 192 5, recognized the dan ger s to ca ne growt h fro m

sa lts of alu minu m and man gan ese and noted that lim e see me d to have little effecton them but found that phosphates gave temporary co rrec tio n . Alt ho ug h he didlittle work with iron, he no ted that ferri c iron would be harmless under existi ngcon di tio ns but warned that ferrous iron " may be ass oc ia te d with th e in fertilit y o fso me o f our poorly aer at ed, highly organic soils." He had noted ea rl ier (16, 1 7, 18 )t ha t soluble "cry st allo id " sa lts of iron , a lum inu m , an d m angan ese were found

CAUSES OF LEAF FREC KLE, LEA F CHLOROS IS, AN D G ROwrH FA ILURE 5

wh er e the soi l pH was bel ow 5.8, th at m an gan ese was not in the soi l so lu tion at pH6.0 or above , an d t ha t iron an d alum inu m at such pH 's exis te d as the hydrosol s offe rric aluminu m hydrat es. He also recogn ized t he possibilit ies of aluminum an dmangane se b eing associa ted with ro o t ro t s bu t poin te d ou t that the so ils invo lvedwe re very low in pho sphorus. In these p ap er s he recogn ized ir on co m pou nds butgenerally see me d to un de re st imate the ir impor tan ce .

Essen tia lly nothing was adde d to thi s a rea of kn owledge in Hawaii for the next35 years. In 19 51, t he plan tatio ns o f C. Brewer an d Co mpa ny began intensive fieldexperi me ntation with coral sto ne an d pho sph orus (P). Amo un ts o f P2Os we re 0 ,200 , an d 4 00 poun ds per acre fro m t re b le superphosphate. Fine ly gro un d co ra lstone was applie d at 4 level s, b eginning with 0 an d ending wit h th e amou nt n eededto bring the p H to 7.0 , as de termined from buffer curves d eveloped by Matsu sak aan d Sherman {1S). As a res u lt, the high est amou n ts in each expe rime n t ran ged fromII tons/acre at Hil o to 17 to ns /acre at Pepeekeo an d 23 tons/acre at Paauhau . In1961 , Clements (2) rep orted on a larg e number of experime n ts dealing with high­level app lications of coral stone, whi ch showe d t ha t the improvement in grow th ofcan e was due to effects of calc iu m (Ca) other than nutrit ion al. When th e pH w aswell b elow 5.6, whe re there was so lub le a lu mi nu m (A I) and iron (Fe) in the so il an dwhe re the mangan ese (Mn) index of sugarcane was above 100 ppm, there was astrong response to ca rbonate, the effect o f wh ich was to red uce the amounts ofthese absorbed by the pl ant. In addition to enhanc eme n t of the Ca conten t of th epl an t , the P level was also increased due t o the imp rovemen t of the rootenvironment brough t ab ou t b y the el imi nat io n of the to xic m aterials.

In 19 61 , Rixon (27) wrote a thesis on studies h e made on t he so ils fro m so m e ofth ese experiments and not ed t ha t , in ge neral, t he high est amo un t of co ra l sto neapplied di d no t ra ise th e pH to 7.0; th at the re was a m arked in crease inexchangeab le Ca , gene ra lly from less than I mcqfl 00 g to as mu ch as 38.82 ; thatthe extractable A l was re duce d alt ho ug h not ver y sharp ly; an d th at even th e highestamo un ts of co ra l stone d id not affect t he cat ion ex cha nge cap ac ity o f any of theso ils.

In 19 61 , Suehisa (30) reported a st udy of the effects of vario us pho sphates andvar ious forms of silicates on the co mp osit ion and yie ld of sudangrass grown o nth ree d ist inc t soi l types : Al u m inous Ferrugino us , Low Humic, an d Hu m ic Latosol s.On ly in t he lattcr was th ere a mar ked increase in yield attributab le to t he silica tes,hence he co ncl uded that t he use of silicates fo r inc reasing P avai lability may b elim i te d . He no ted , however , that so lub le silicates h ad very pronou nced effe c ts o nreducing ex t ractab le AI.

In 19 6 2, Monteith (20) wr ote a th esis co m p ar ing t he effects of calc iu mca rbonate wi th calcium mctasili cate on thc growth of sudangrass on a Ferrug ino us(Pu hi) and a Hy drol Humi c Lato sol (Aka ka) . He ob taine d increases in yicld forbo t h in t hc latte r soi l, but on ly for t he silicate in thc former.

As a resul t o f these st ud ies, G . D. Sherman (20) pl anned a field ex pe riment usingTVA slag an d phospha te a t Puh i, and the senior author (3, 4 ) p lanne d one wit hso di um silica te, coral stone, and superphosphate a t Kilauea, both being inlow-productivity soi l areas on Ka uai, and bo t h experiments were ins talled in 19 60.T he Puhi experime n t showed excelle nt yi eld responses to the TVA slag . The

6 HAWAII AGRICULTURAL EXPERIMENT STATION

Kilauea experiment showed a significant decrease for coral stone, no response tophosphate, and a significant increase for the sodium silicate. There followednumerous field experiments using either TVA slag or the considerably betterpseudowollastonite (CaSi03 ) , developed as a result of cooperative efforts by thesenior author and officials of the Hawaiian Cement Company. Efforts to explainthe increased yields resulting from the silicate usually stressed decreased Al andincreased P and Ca. In 1967, Clements, Putman, and Wilson (6) published theirfindings on studies involving the very toxic soil at Kilauea, Kauai, and demonstratedclearly for the first time that ferrous iron (Fe++) was by far the most dominantconstituent of the soil solution, causing reduced root growth, reduced yield, and aleaf freckling.

Other work includes that by Mahilum (13) on the coral stone experimental plotsoils . His thesis verified the finding reported above relating to the lowering of the Allevels by lime and also pointed to the readiness of leaching of Ca. He observedparticles of coral stone still remaining after several years and showed that, by raisingthe temperature of the soil, these disappeared and the pH rose to the anticipatedlevel.

Fox (8) and his associates published data showing the distribution of silicon (Si)within the cane plant and also the seasonal variation in the absorption of silica(Si02 ) . They emphasized the importance and activity of soluble Si.

Elsewhere, Halais and Parish (10), in Mauritius, made use of the manganese: silicaratios in diagnosing silicate needs of cane. Even though Mn seems less importantthan other elements on potentially troublesome soils, a high ratio warns of toxicconditions. Thus, Halais (9) has also been able with the use of silicates to regeneratesenile soils where Fe has been a dominant factor.

Samuels and Alexander (28), in Puerto Rico, have confirmed results obtained inHawaii and are extending their research to include any possible roles Si may have insugarcane physio logy, particularly in the various enzyme systems, including thoseassociated with photosynthesis and storage of sucrose.

Fundamental research continues in Japan on the role of Si in rice, begun in 1936by Okawa (23). The effect of silicate on the structure of the rice plant and itsresistance to insect and fungus attacks implies that Si, though not an essentialelement in the botanical sense, is a valuable agronomic material and the cheapestavailable to accomplish the purposes discussed.

A better understanding of the Fe complexes in soils results from the recentreview by Oades (22). Organic materials have a marked effect on the formation ofFe++ in soils having various redox potentials. Motomura (21) shows that a soil maybe quite toxic under one set of conditions and not under others, as is theexperience in sugarcane fields in Hawaii. The fundamental approach being made atthe International Rice Research Institute (26) to portray the multip licity ofchanges occurring in soils as they become flooded or drained has application farbeyond the rice paddy.

It has long been recognized that ammoniacal fertilizers leave acid residues. Thus,Pierre (24, 25) reported that the acidity which 1 ton of ammonium sulfate (21.1percent N) adds to the soil requires 2249 pounds of calcium carbonate forneutralization, one ton of ammonium chloride (26.1 percent N) 2786 pounds, etc.

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE 7

Kanehiro and Hadano , (11) reported t h at , per pound, am mon iu m nitrogen ,amm o n ium su lfa te , am mo ni um nitra te , m ono-ammonium phosphate, and di arn ­m onium phosphate all have abou t the same effec t on soi l aci d i ty. These forms ofnitrog en are com monly in us e becau se of lower cost, bu t rarely does one find inHawaii a sus ta ine d program of ad ding neutrali zing agents. Samuels an d Gonzales(29) more rec en t ly showed the long-range effect of co n tinue d use o f sulfate ofammonia to be a very serious ac cumulation of soil acids.

SUGARCANE LEAF FRECKLEThe term "leaf fre ckle" h as been applied to maladies of more than one origin . As

used in this bulletin, it applies to a condition which manifests itself on sugarcanegrowing on aci d soils, particularly following the winter period (Fig. I and 2). It hasbeen especially severe on poorly drained Aluminous Ferruginous, Humic, an dHydrol Humic Latosols following cold periods of heavy rainfall. Fields of plantcane planted during the March to July period generally develop vigorously, and byDecember the crop appea rs to be gro wing extremely well. As the rains increase an dthe temper ature drops, however, growt h slows d own and the leaves, including thesp ind le leaf , begin to show sma ll elongate spots (F ig. 2a) whi ch fail t o developch loro p hy ll and hen ce appear as ye llo wis h areas . As the leaf ages, the spo ts be comereddish (Fig. 2b and c), then brown, and fina lly dark gray brown and ne crotic . Inthe m eantime, the sp o ts enlarge , p re vious ly green areas lose th eir chlo rop hy ll , andthe necrosis spreads (Fig. Ib) . Fre que n t ly, mir rorlike, silvery spo ts develop on thedi st al surfaces, part icularly those ex posed to fu ll sun ligh t. During th e time wh en thespots are re ddi sh, t he f ield may have a b ro nze cas t (Fig. la), and later, as ne cr osissp reads, the older leaves di e prematurely. In con trast to healthy plan ts , which m ayhave 10 to 14 living leaves, aff ected pl ants may have on ly 7 or 8, but even theseleaves are developing freckles an d h en ce are co ns ide ra bly less than fully efficie n t.

Because of th e fre ckles and the co ld weather , growt h slows d own co nsiderab ly .Eve n when spring co m es, growt h do es not res ume as it no rmally sho u ld , an d whaton ce see me d like a fine cro p ta kes on an appearan ce of openness wi th small can etops and slender upp er st a lks . Ofte n weed s begin to grow , an d at har vest th e ca nesare light in weight and the y ie lds ver y dis appointing.

So il conditions associat ed wi t h this deve lo p men t include poor drain age an d lowpH , usu all y in the ran ge of 4. 0 to 5 .0 . Profiles of t he soi l a lmost always showoccasion al dark masses whi ch sme ll of hydrogen su lfi de .

Experimental BackgroundT he first convincing de monstra tio n of the correction o f fr eck ling res u lted from

fie ld experiments at Ki lauea and Puhi , both on th e Island of Kauai, fo llowing th euse o f TVA slag on th e A luminous Ferruginous Lat o so!. Lat e r , sim ilarly strikingdemonstration s o cc urred on th e very ac id upper fields a t Paauh au, Island o f Haw aii,fo llowing th e us e of a re lativel y pure calciu m met as ilicat e. Wher eas th e TVA slagwas sho wn to be an amorphous geh leni te (Ca 2 AI2Si0 7 ) with some calc iu mmetas ilicate and residual p hosphate , th e new product is a high -qu ality , crys ta llinepseudowollastonite (CaSi0 3 ) wit h so me ra nkinit e (Ca3 Si2°7 ) included . Similardemonstrations were seen on th e upper, poorly d rained Hydrol Humic Lat o sols a t

8 IIA \\'A II AGRIC ULTURAL EXPER I~IENT STATIO1\:

Fig. I. Su gar cane leaves showing bronzing and fre ckli ng. Top (a) shows the redd ish b ro nzingof o ld leav es as well as the fre ck ling of young o nes (1153·263). T he dark green lea f was detach edfro m the st a lk and inser ted for co nt ras t. Bottom (b) sh ows, from left to righ t , vario us stages ofleaf fr eck ling with two normal leaves ins erte d fo r co n t ra st (H5 1·22 79) .

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE 9

Fig. 2. Sug arcane leaves sh ow ing increasing int ensity of frec kl ing from a through h (H53-26 3).Low er left and low er righ t (e and h) sho w lea f underside. Lower ce nte r if and g ) sh ow s views ofuppersi de of leaf .

Ta

ble

2.G

row

than

dsi

lica

anal

ysi

so

fsu

dan

gras

son

vari

ou

ssi

lica

tem

iner

als

Sil

ica-

perc

en

tdr

yw

eigh

tD

rym

atte

r-g

ram

sp

erpo

t

(to

ns

of

mat

eria

lpe

rac

re)

(ton

so

fm

ater

ial

per

acre

)

Pred

om

inan

tM

iner

alch

em

ical

02

48

02

48

Pse

ud

owol

last

on

ite

CaS

i 03

.28

2.1

95.

136.

4010

.136

.23

1.4

19.6

Ran

kin

ite

Ca

3Si 2

07

.28

2.3

14.

214

.56

10.1

41.5

48.

536

.0L

arni

teC

a2S

i04

.28

3.16

2.9

45.

7510

.132

.044

.519

.1A

nort

hite

CaA

l2S

i20

S.2

80.

400

.80

0.8

310

.128

.419

.120

.5G

ehle

nite

Ca2

Al2

Si0

7.2

81.

382.

322.

6010

.136

.820

.832

.8T

VA

slag

a(a

mor

phou

s)C

a2A

l 2S

i07

.28

1.4

51.

852.

08

10.1

38.9

42.6

42.

4T

VA

slag

b(c

ryst

alli

ne)

Ca2

Ah

Si0

7.2

81.

482.

51

2.76

10.1

40.1

48.5

43.

6H

awai

ian

Cal

ciu

mM

etas

ilica

teC

aSi0

3.2

82.

49

3.42

3.28

10.1

45.5

44.4

46.

5

aTh

isT

VA

slag

was

amo

rpho

usan

dw

asd

eter

min

edto

be

geh

len

ite

wit

hso

me

calc

ium

met

asili

cate

and

also

som

ere

sidu

alp

ho

sph

ate.

bIl

eca

use

earl

yin

this

wo

rkth

eam

orp

hous

nat

ure

of

the

TV

Asl

agw

asth

ou

gh

tto

be

imp

ort

an

tin

its

reac

tivi

ty,

aq

uan

tity

wa

sta

ken

,h

eate

dto

am

olt

enst

ate

,an

dth

enco

ole

d,p

rod

uci

nga

com

ple

tely

cry

stal

line

pro

du

ct.

o ::r: >- ~ >­ o ::0 n c:: r­ ..., c:: ::0 >­ r- t'l >< ""t'l ::0 §: t'l Z ..., rJJ ..., >­ ..., 23 z

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE I I

Pepeekeo, Hakalau, and Onomea. Similar experiments a t Pahala and Naalehu, wherethe Hydrol Humic Latosol s ar e mu ch less acid and better drained, even though co ld ,and wh ere ti ssue silica level s are high, failed to yield any growth response to silica te .Fre ckling at these pl antations is rarely seen . At Olokele Sugar Company an d atWailuku Sugar Company, (on the Island of Maui), where the warm Low HumicLatosols have a generally higher pH, dr ainage is good , and ti ssue silica is moderatelyhigh, no response has been noted nor is there any fre ckling.

Useful Silicates and Local Silicate MaterialsConsiderable time was given to a search for other form s of silicate which might

be useful to the plantat ions. Small am ou n ts of various silicate minerals, made fromre agent grade chemicals by Ri chard Berger of the America n Cement Company, weretested, using sudangra ss in no . 10 cans and the known toxic soil from Kilauea. Thedata (6) are in T able 2 and should be regarded as only partially quantitative becausethe degree of fineness of the minerals varied, although all were very fine.

That several calciu m silicate minerals might be used is evident. Anorthite,considere d to be CaAI 2Si2°8 , is rel atively unavailable to the plant and h ad leasteff ec t on the growth of suda ngrass . Yet gehlenite , co nside re d to be Ca2AI2Si07 ,

was cons ide rab ly more avail able and did improve growth rel atively well. In general,the ca lc iu m sili cates provided more re adily avail able silica and also improved growththe most. The amorphous nature of TVA slag is not essential.

Various local materials were similarly tested and several were quite good. Someof these data are summarized in Table 3. Kahili rock cam e from Crater Hill quarrynear Kilauea, Kauai, and is a soft ro ck used for plantation road surfacing. Bouldersare spread over the ro ad and he avy bulldozers go back and forth over them ,crushing and co mp ac t ing them into a smooth road sur fac e. Prior to the recentwork, at least two of the pl antation m an agers noted an apparen t improvement inthe appeara nce o f the cane near th e road. The tr ach yte ro ck s ca me from the Islandof Oahu, the volcanite from Puuwaaw aa on the Island of Haw aii. With each o fthese, there was some growth improvement, altho ugh none compared with th e

Table 3. Silica content (percent Si02) of sudangrass grown on soil mixed with various mat erials

Tons of material per acre

Material 0 2 4 8 16

Kahili ro ck .34 .68 .89 1.62Kahili rock + 40% cement .34 .94 1.45 2.01 2.32Trachyt e + 40% ceme nt .34 1.10 1.69 2.04 3.93Volc anite + 40% ceme nt .34 .75 1.26 1.89 2.53Cement .34 1.47 1.53 3.6 7 2.69TVA slag .34 1.45 1.85 2.08 2.82Hawaiian Calcium Metasilicat e .34 2.49 3.42 3.28 4. 25

12 HAWAII AGRICULTURAL EXPERIMENT STATI ON

Hawaii an Calcium Metasil icate. Ordinary building ce ment by its elf gave some goodresults but, b ecause of its high CaO con te nt , chlorosis an d lea f burn often resulted.A red cinde r from the Isl and o f Mol okai and blue basalt from Oahu, when veryfine ly gro un d , ga ve so me sm all imp rovem en t although not co nsiste n tly .

Thus, there are sev er al materials which might be used. Where these were used intrial s for correc t ion of frec kling , sugarcan e sh owed improved growth but this neverco mpare d well wi th sugar can e tr eated with the ca lci u m m et asili cate (made by theHawaiian Ce me n t Co mpan y) nor were the benefits as long lasting . The ir product,cu rren t ly in use as a un iformly groun d powder, abo u t 200 mesh, is prepared byburning a mixture of h igh -quality local coral stone an d im ported sili ca sand.

The Causes of Freckling in SugarcaneT he ne xt problem needing clarificatio n is the determin ati on of the cause or

ca uses of frec kling and poor gro w t h . Several aven ues wer e fo llo wed: analysis ofaffec te d vs . normal leav es, roo t analys is , soil cu lture studies, and soil extractanalysis.

Analysis of Cane LeavesKil au ea Plantatio n Co mpany installe d a co mp licate d experiment in one of it s

wo rs t freckling fields and used 3 var ie t ies , TV A slag grou nd to - 16 and -60 mesh ,and ap plie d a t 3 and 6 tons, respectively , an d Mol okai cin de r gro un d to - 180 m esh,eac h appl ied b roadcast and wo rked into t he soi l vs. applicat ion under the seed a tpl anting. T he experiment demonstrated clearly tha t the cin der was not effective,t hat broadcasting and incorp orating t he slag into the so il with a ro tovato r wa s veryeff ective , that pl acing it under the seed ess entially nullified its influence, that t h efine r t he slag was grou nd the greate r wa s its effec t ivene ss, and that the largeramount was clearl y superior to the smaller amou nt wh ich, in turn, was superior tothe chec k.

Beca use the ex periment involved 96 pl ots an d becau se so m an y answe rs werealready ver y evide nt long before harves t, th e pl an tation decided to h arvest on lyselected pl o ts fo r expe rime n ta l data. Regular cro p log sam p les wer e taken from thera toon whi ch fo llo we d th e harves t ed pl ots . In add ition t o the regul ar inde x tissues ,sampl es wer e take n from the yo ung bl ad es (3 , 4, 5 , an d 6) an d the o ld bl ades (7, 8,9, et c.). T hese wer e di vided into bottom , mi ddle, and top th irds an d analyzed . T heresulting dat a wer e put t hro ug h co mp uters progr ammed fo r mult iple st epwiseregressio n . T he res ults for H 39 -7028 on ly are repo rted in Table 4, which lists thesimple co rrela t ions be tween the particu lar element in top , m idd le, and bo ttomportions o f young leaves (3 , 4, 5, 6) an d of o ld leaves (7 , 8, 9, e tc.) wi th to nspo l/ acre. In ge neral, Ca and Si0 2 are always posit ively correlated with y ield , whil eFe, AI, Mn , Co , and T i arc ne gat ive ly corre la ted . Sin ce t he fre ckling is inverse lyrel at ed to yie ld , it is reasonab le to con clude that both t h e accumula tio n of Fe, AI,Mn , e tc. , an d the lo w levels of Ca and Si0 2 in t he leav es are t he assoc ia te d factors .The column headed R 2 re pre sents the sum of th e products of th e correlations an dthe st andard partial regressions, whi ch are very high . With 95 to 99 percent of th evari abi lity account ed for , th e factors listed include at least the most troub lesomeon es. T he co lu mn at the righ t ind icates t he order of im po rt an ce as de termined by

Tab

le4.

Cor

rela

tion

sb

etw

een

the

ash

com

po

nen

tso

fth

ele

aves

and

tons

po

l/ac

reyi

elds

(H39

-70

28)

(r=

.76

5fo

r1%

,.6

32

for

5%)

Lea

fpo

rtio

nC

aS

i02

Co

Fe

Mn

Al

Ti

R2

Ste

pwis

eo

rder

Yo

ung

leav

esT

op

.79

4.8

42

-.20

2-.

260

-.26

1-.

359

-.0

327

.987

8**

Si0

2,

Fe,

Ca,

Co,

Ti

Mid

dle

.766

.926

-.3

30-.

32

8-.

339

-.58

3.2

71.9

93

2*

*S

i02

,Fe

,T

i3,

4,5

,6B

ott

om

.69

7.9

39-.

590

-.30

7-.

480

-.55

5-.

170

.990

4**

Si0

2,

Fe,

Ca,

Al

·)Id

leav

esT

op

.807

.927

--.

755

-.25

7-.

756

.97

10**

Si0

2,

Fe,

Al

7,8,

9,et

c.M

iddl

e.8

54.8

47

.28

7-.

285

-.34

9-.

604

-.06

5.9

478

**C

a,S

i02

Bo

tto

m.9

04

.930

+.0

11-.

54

8-.

583

-.5

74-.

364

.969

5*

*Si

02

,C

o,C

a

o > c: Vl tTl

tr: o "'l r­ tTl > "'l

"'l

::0 tTl

o ;;>:: r­ ,.tTl r­ tTl > "'l o :::: r­ o ::0 o tr: .en > z o o ::0 o :E ...., :::: "'l > t"" c: ::0 tTl

ee

14 HAWAII AGRICULTURAL EXPERIMENT STATION

the stepwise regression . After completing the entire run for a particular set of data,the computer printed out the most important of the 7 X's used first, then thesecond, and so on, until no further significant improvement in the R 2 wasobtainable. Thus, it can be seen that the dominant factors are the actual Si0 2

content of the leaves, a positive factor, and the Fe content of the leaves, thedominant negative factor.

The second variety (H53-263) used in this study showed quite similar results,with Si02 being the dominant factor in each case and Fe, AI, and B being thesecond or third in each case. Ca appeared only once and this as a third factor. Thethird variety (H51-2279), however, was quite different (Table 5). Although ingeneral the R 2 's are of the same order and sign, they are much lower and thestepwise order is very different. Where H39-7028 showed Si0 2 to be the dominantfactor in five of six runs, and where H53-263 showed it to be dominant in all sixruns, H51-2279 shows it only once as the first factor and once as a second. Thereseems little order to the results . This is especially pertinent since this variety hasfrequently given disappointing results in fields where silicate has been applied.Perhaps one reason for this is that H51-2279 seems less able to absorb Si02 thanthe other two . Untreated plots for the 3 varieties show 0.52 percent for H39-7028,0.25 percent for H53-263, and 0 .24 percent for H51-2279. The treated plots forthe same varieties show 1.41, 1.37, and 1.10, respectively. The higher level forH39-7028 perhaps accounts, in part at least, for its having been a standard in thetoxic soil area prior to the silicate era.

When the actual leaf composition values are displayed in Table 6, thecomparisons for Fe do not seem to be very striking. Ca and Si02 levels aremarkedly different, however, and Mn also is markedly lower in the treated plots,although the actual levels even for the checks are not very high when comparedwith other areas. AI also is lowered by treatment, even though the levels are notvery high compared with levels found elsewhere. Thus, it is possible that, once theMn, Fe, and Al get into the leaves, if they do not encounter certain minimum levelsof Si02 and/or Ca, they localize in such concentration that they cause necrosis. Yet,the same levels of these toxic materials are prevented from localizing by the presenceof Si02 and/or Ca, as Williams and Vlamis (35) have noted for Mn in barley leaves.

An alternate interpretation is that the Fe, AI, Mn, etc ., in the soil solutionbecome very toxic to the roots, and that this is the primary cause for the yielddecline, with freckling an inevitable consequence, compounding the difficulties forthe plants. Many studies along these lines were undertaken in an attempt toassociate composition with the toxicity shown by freckling. A new sampling of thetop thirds of old leaves taken from the same field as reported above was made fromcheck plots (X) as well as the 6 tons of 60-mesh slag-treated plots (D). These werethoroughly washed and, to avoid contamination in grinding, were cut into piecesabout 1 inch square and ashed without grinding. The ash was taken up completelyand analyzed by atomic absorption. To add more relative meaning to the search,similar samples of deep-green leaves showing no trace of freckling were taken fromcomplete nutrient cultures being grown in silica sand (Z). These data are shown inTable 7. Comparing the freckled leaves from plots X with the slag-treated leavesfrom plots D, there was more lead (Pb), Zn, AI, Fe, and Mn than in the leaves from

Tab

le5.

Cor

rela

tion

sb

etw

een

the

ash

com

po

nen

tso

fth

ele

aves

and

ton

sp

ol/

acre

yie

lds

(H51

-227

9)(r

=.7

65fo

r1%

,.63

2fo

r5%

)I~ c en

Lea

fp

ort

ion

Ca

Si 0

2B

FeM

nA

IR

2St

epw

ise

ord

er1

1h 0 "1 r-T

op.6

48.6

33-.

25

5-.

208

-.6

06

-.30

2.9

37**

Ca,

AI,

Fetr

l

Yo

un

gle

aves

;..M

iddl

e.5

27.5

34- .

504

-.24

3-.

321

-.11

6.8

27S

i02

,B

,C

a"1

3,4

,5,6

"1B

ott

om

.648

.575

-.11

9-.

67

6-.

51

4-.

281

.954

**

Fe,

Ca

:<l

To

p.7

38.6

00-.

522

-.0

71

-.5

46

-.22

8.7

32

Ca,

Si0

2tr

l

Old

lea

ves

oM

iddl

e.8

00.6

58-.

46

6-.

420

-.59

2-.

370

.900

**

Ca,

Fe~

7,8,

9,et

c.r-

Bo

tto

m.7

02

.595

-.55

5-.

431

-.73

8-.

40

6.6

24M

n,B

..trl r- trl ;.. "1 o ::r: r- 0 :<l 0 en Iii ;..

Tab

le6.

Eff

ect

of

TV

Asl

ago

nyi

eld

and

onco

mp

osi

tion

of

can

ele

aves

of

the

thre

eva

riet

ies

Z 0 o :<l

Var

iety

To

ns

Ca

Si0

2M

gC

oF

eM

nA

lT

iZ

nM

oB

I~po

l/ac

re::r: "1 ~

H39

-702

8T

reat

ed18

.4.6

211.

421.

1.6

142

8550

928

.183

11.9

I~C

hec

ks

13.9

.451

.52

.91.

613

612

856

636

.180

13.3

H53

-263

Tre

ated

18.1

.485

1.40

1.8

1.1

116

8246

723

.209

14.4

Che

cks

15.2

.276

.25

.7.7

125

126

605

28.2

252

1.2

H51

-227

9T

reat

ed18

.2.4

921.

17.6

4.3

126

8860

927

.204

13.9

Ch

eck

s15

.5.2

86

.24

.6.3

155

150

9010

23.1

93

16.8

'"

16 HAWAII AGRICULTURAL EXPERIMENT STATION

Tahle 7. Composition of top thirds of leaves (ppm) from severely freckled, silicate-treated,and nutrient culture plants (H53-263)

Elementor compound

PhZnCuCrAlFeMnCaSi02

Source of material

Freckled Silicate-treated Nutrient cu ltur e(X) (D) (Z)

4.3 2.7 1.34.6 3.7 4.6

.9 1.3 .9

.15 .11 .2418.0 8.0 55.078.2 50.8 171.088.2 34.8 119.0

1,850.0 6,500.0 16,500.02,300.0 65,000.0 8,600.0

plots D, but much less Ca and Si02 • The perfectly green leaves of the nutrientculture plants, which turn a clean, uniformly light tan color when they die on thestalk, had much higher contents of the toxic elements than leaves from plots X, butvery much higher Ca and Si02 levels. The same conclusion was reached as afterTables 5 and 6; namely, that leaves which are low in Ca and Si02 become moresubject to the localization of the toxic elements and, even though the levels of theseelements may become high, if Ca and Si02 are abundant, the toxic effects are notrealized.

Analysis of RootsIf the toxicity exists in the soil solution, then root analysis may yield some

cogent data. In Table 8, roots taken from the severely freckled field plants andthoroughly washed were compared with those taken from plants grown in acomplete silica sand nutrient culture.

Roots from severely freckled plants are very much higher in Fe, AI, Mn, copper(Cu), boron (B), and Zn, and very much lower in Si02 , Ca, S, Mg, and P. Thus, thesame general pattern continues to emerge.

Soil Solution ToxicityA soil culture series was set up in no . 10 cans, using sudangrass and a 15-15 -15

garden fertilizer, which was a mixture of ammonium sulfate, diammoniumphosphate, and muriate of potash. The data shown in Table 2 are typical of theseruns. In this specific case, when the plant crop was harvested and the culturesfertilized again, ratoons of the silicate-treated pots generally grew very well but thechecks were complete failures. Extracts of this soil were prepared to identify thetoxic material. Early in this work, soil and water were mixed on the basis of I to I .

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FA ILURE

Tab le 8. Composition of ro ots of severely fre ckl ed plants co mpa red with t hose grownin nutrient culture (H53-26 3)

17

Elem entor co mpo u nd

KP

CaMg

SSi0 2

MnZnBCuAlMoFe

Roots fromnutrient cu lture (Z)

Percent dry matter

.90

.07 6

.63

.22

.58 32.00 (est im at e)

Parts per million

15.036.0

6.649 .0

150 .06.7

288.0

Roots fromfreckled plants (X)

3.3 8.05 4.044.088.119.99

25 1.0200.0

3 1.066.0

1,540.02.2 8

12,406.0

Gen erally, the mixture was allowed to stand for as long as 2 weeks with occasion alshaking. Later , a week o f soaking was followed by 24 hours a t 85 C in an o ven . Stilllater , the mi xture was allowed to stand for 24 hours an d then was au to claved a t 15psi for 1 hour an d ce nt ri fuge d until clear , thus maintaining th e natural pH.

Two samples of Ki lau ea soil were used: on e in which r atoon growth w as aco mp lete failure (used), the other was the soil as received from the fi eld (n ew) .First , a wh eat -seed ge rmi na tio n was run on the extract. Dist illed wat er was used as acheck . The " new" ext rac t a lso was co ncen tra te d five times to imitat e the amo untof moisture at normal field ca pacity . The germination res u lts are in T able 9.

Ta b le 9. To xicity assay of Kilau ea soi l ex trac ts (whea t )

Percen tExtrac t germ ina tion

Distill ed H 20, chec k 100" New" ex trac t 4 7" Used " ext rac t 2"N ew" co nce ntrated five ti mes 10

Sh ootgrowth

strongpoornilnil

Ro otgro wth

strongnilnilnil

18 HAWAII AG RICU LTU RA L EX PERIMENT ST ATION

T wo points are cle ar. T he toxici ty ex ists in t he so il as received bu t is in te ns ifiedfo llowing us e of the 15-1 5-1 5 fertilizer mi xture . I t is clear t ha t th e to x ici ty in thefie ld is the result of natural weathering, since soi ls whi ch h ave n ever been fe rti lize dare also toxic, but the use of ammoniacal fertilizer s great ly intensifie s the p roblem.

Scanning and Resin Exchange ColumnsT o ide nt ify the ac tual toxi c agent , tw o procedures were followed: t he firs t

involved scan n ing the ex t ract in the atomic absorp t ion spectro p ho to meter; thesecon d invo lved resin excha nge co lu m ns an d chem ical methods fo r the separationand identification o f the co mpon en t mat erials, t esting each fr act ion for toxicitywit h wh eat-seed ge rmina ti on .

Sc an ning the soil extract showe d the presen ce of man y po ssibl e toxic materials:Fe , Mn, A I, Ti, Ni , Co, Cr , Pb, Cu , and Zn . T he " use d" extrac t had 22 0 .0 ppm ofFe and 10.0 ppm Mn, while the " new ," which also was toxic to wh ea t ge rm ina tionthough less so , showe d 20. 0 an d 1. 85 ppm, resp ectively . Ch emical test s sho we d t heFe to be mo stly Fe++ . One analy sis sh owed 0 .33 mm ol e/lOOcc Fe+ + and 0.07mmole/ 100cc Fe+++. T he buffer curse of the so il extrac t was shown to be similarto that of Fe+ +.

Chro ma tograp hic analysis of the ex trac t showed considerab le organ ic m aterial,with ami no aci ds making up 30 to 4 0 ppm o f the total ex trac t. Am ino acidsidentified wer e alan ine, glycine, serine, valine, t hre oni ne, asparti c aci d , leu cine , betaalani ne , tyrosine , gamma ami no butyric ac id, phenyl a lani ne , lysine, argi n ine ,histidine, and glu tamic ac id .

Ion exc ha nge techniques wer e employed to o btain 4 fr action s: a neutralfract ion, a weak base ca tion fraction con tain ing amino aci ds, an aci dic fr act ionincluding ino rga n ic an ions and organic aci ds, an d a strongly ab sorbed ca tionfr action. None of t he first 3 fract ions sho wed to xicities to wheat ge rm ina tio n . T hefou rth fr action was eluted with 2N HCI, and the eluate was m ade alka line wi th 6NNaO H to separate the metall ic hydro xides. A co pious blue-green precipitateappeared , was fi lte red off, wash ed free of alka li, di ssol ved in a minim al amoun t o fdilute HN0 3 , an d the pH adj us te d to 3 .9 5 with KOH (the pH o f the " use d" soi lex t rac t ). As was soon appare n t from the co lo r and in st ability o f the resul tantso lu tion , the Fe++ had been ox idize d to the inso lub le Fe (OHh , and the so lu tionwas non to x ic to wh ea t ger mina tio n .

When , however, the HCI in the cat ion eluate was removed b y di stillation underreduced pressure, the co mposit ion of the so lu tio n of the crys talline residue wasesti mate d to be 0.004M FeS04 and 0.0 23M NH 4CI a t a pH o f 3 .95 . When aso lu tion of FeS04 half this stre ngt h (0.00 2M) an d with 0.02 3M NH 4 CI was m ad eup , adj us te d to pH 3.95, germ ina t ion was a co m p le te fa ilu re with the swo lle n see dssho wing the same darken ing as had been noted whe n ge rm ina tio n was a ttempted inthe o rigina l soil ex t ra c t. T he iron sulfate used alo ne in ano the r trial showed that ita lo ne was the toxic age nt. Of special in teres t is the fa ct that Fe(OH h is so lub le inweakly acidic so lut ions and th at sili cic aci d is insoluble in them . Thus, a t Kilauea, inits ac id Alu m inous Ferrug inous Lato sols , Fe++ undoubtedly is th e dominant so ilpo ison . As it is di ssolved off th e so il miner al s , it acc u mu lates in the so il so lu tionand, becaus e o f po or ae ra t io n during the heavy rainfall period, remains in the

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE 19

ferrous form. Under summer co nd it io ns with good aer ation , th e same soil would b eless toxic be cause ferrous iron would be come ferri c and would be precipitated.

At Kil auea, Al undoubtedly is also a part of the problem be cause the affectedlands are potential bauxite sources . In soils at Akaka as well as Puhi, Monteith an dSherman (20) have shown that extractable Al is markedly reduced by both lim e andsilica te. Suehisa et al. (29) have shown striking reductions of extractable Alfollowing treatment with eit he r phosphat e or sili cate . Cl em ents (2, 4 ) ha s sho wn amarked reduction of plant mangan ese and root alumi nu m as a result of liming witheither carb ona te or sili cate .

Monteith and Sherma n sugges ted also, as have others, that one of the principalroles of the silicate is to rel ease phosphate to the crop . If th er e is any co n nec t ion atall, it is doubtful that this is a principal fun ction, sin ce the same in crease in pl antphosphates follows ca lciu m carbo nate application. Thus, in Table 10 , am p lifie dphosphorus index (API) readings for sugarcane ar e reported for an "amounts ofphosphate" by "amounts of co ral stone" e xp er iment and show a larg er in crease inplant phosphorus from co ral than from phosphate.

T abl e 10. Plant (API) phosphorus readings for sug arcane in a phosph at e X coral sto neexpe riment (Hil o)

Pounds of 1'205 per acre

Tons of co ra l stoneper ac re o 200 400

Average 1492

aAmplified phosphorus index.

o2.05.5

I !.O

1429167114101459

1387 14801762 16741975 20282107 1972

1808 1792

1428170118031853

In Tabl e 11, API readings are shown for a silicate exp erime n t , also a t Hilo. Aswas the case with coral stone , ca lciu m metasilicate gave a larger in crease inphosphate uptake than did the phosphat e fertilizer. One interpretation of thisincrease, due to either carbo na te or sili cate , has been th at th e treatment affe c ts th eanion ex change complex. It is much more likely that th e increased absorption is theresult of improving the root enviro nmen t b y precipitating to xic elements, and rootgrowth is greatly stimulated (see later), re sulting in a greater uptake of nutrients ingeneral and phosphate in particular. It is a lso very doubtful that the in crease ingrowth shown by Monteith and Sherman w as due to the phosphate effect , since thecarb o nate had the same effect on phosphate uptake but not on growth . Although aswill be noted for Ke aau, growth improvement was effec te d equally, whetherCaC03 , CaSi03 , or MgC0 3 was used. At Kilauea, Paauhau, Pepeekeo, et c ., on landswhere fr eckling occurred, the response to the silicate was more decisive than to the

20 HAWAII AGRICULTURAL EXPERIMENT STATION

Table 11. Plant phos phoru s (API) readings in a phosphate X silicate experiment (Hilo)

Tons of calcium met asilieate per acre

Pounds ofP20S per acre 0 2 4 6 APIa

0 2093 3080 2720 2777 2668400 1788 2840 3196 3348 2817800 2819 3287 3382 3174 3190

1200 2804 3380 3674 3645 3376

Average 2376 3171 3264 3246

aAmplified phosphorus index.

ca rbo na te. Table 12 shows f iel d yi elds in A kaka soils for plant as w ell as ratooncrops of sugarca ne. In both plant and ra too n crop s, the silicate outyielded theca rb onate t reatments.

Although the ca rb o n ate treatment in the pl ant cro p outyielded the c indertreatment, which wa s wi thout effect, in the ratoon the difference w as just short ofth e 5 percent level. Similar results h ave been o b t ained at other pl antations .So me how, eit her the ac tion of the sili cat e d iffers fr o m that of the carb onate , or theresulting products differ in st ability . If the act io n is one of precipitating such to xi cel ements as Fe++ , AI+++, Zn++ , Ni++, etc., it is not unlikely that th e sili cates of

t hese elem e n t s are mo re resistant t o weatheri ng breakdown than ar e th e carb on ate s.

Table 12. Tons o f cane per acre-Akaka (average of treated plots)

Supp lement Plant Rat oon

TVA slag 86 143Olivine cinder 73 130Cora l stone 80 134

usn 4.3 4.1

Specificity of the Silicate ActionSince it now seems clear that the ac t io n o f calciu m metasilicat e is so m ew ha t

different from that of calci u m carb o nate, the question is whether so me specificcharac te r o f ca lciu m metasili cat e is inv olved. An array of an io ns was sel ected whichwould be so me w ha t co mparab le t o sili cate in the ins o lub ility of co mpo u nds withsuch ca t ions as Fe ++, AI+ ++, Z n++, Mn++, Ti++, e tc. The results are summarized in

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE 21

T able 13. This effort resulted in part al so from an o b se rva tio n test in which the

toxic soil wa s treated wi th very large amounts (1000 pounds per acre) of various

eleme n t s including the minor elements. The response to sodium m ol ybdate w as sooutstanding that it suggested that fre ckling was m erel y a symptom of a ve ry severe

molybdenum defici ency. This id e a was di smissed when the d ata b e came ava ilab le .The molybdate anion, lik e phospho-tungstate an d others, including silica te , ca u ses

much the same kind of insoluble precipitate with the known possible toxic cat io n s.

These d ata, so far a s these experiments go , al so eliminate the id ea that silico n is a n

essen t ial element and that fre ckling is a silicon-deficiency symptom.It is likely that the effec t ive n ess o f the very so lu b le so d iu m sili cate is due in p art

to it s strong alkalin it y, an d that the more or les s in soluble calc iu m m etasilicate is

effect ive b e cause , a t it s surface, the pH is est imated t o b e so m ew hat above 11 . I t isa lso p o ssible that the calciu m silicate wh ich d issolves in the so il so lu t io n reac ts with

the soil particles which con t ai n the toxi c ferrous iron, thu s in sulating the p lan t

roots from co n tact with it.

Table 13. Dry matter yie ld of suda ngrass (grams per pot) (Trea tment range: 0. 25 to 2.0 ton sper acre silica te equivalent , exce pt for ammo nium ph osph om olybdat e)

Mat erial

Hawaiian Calcium Mct asilicat eAmmo nium phosph om olybdat e (2-1 2 tons/ac re]Sodium silica teAmmo nium phosph o-tungstat eTVA slagSodium molybdatcSodium pyrophosph at cCalcium carbo nateSodium te llura teAmm oni um paramolybdat eRankinite no. 2GypsumChecksSodium arsena te

Rating

I23456789

10I I12

13

Dry wcight(average)

23.423. 122.322.021.421.320.720.220.219.919.819.218.918.5

Effects of Calcium Silicate vs. Silicic Acidand of Fertilizers on Soil Toxicity

The soil toxicity reported fo r Kil auea re sults from natural weathering, since

lands whi ch never hav e be en fertilized give st ro ng responses to calc iu m m etasilicate .Howe ver, it is a lso cl ear that am mon iacal fertilizers, b e cause of the a cid re sidue s

they leave , ad d to the am o u n ts of Fe+ + , AI, e tc., in the so il solution , while a t t he sa me

time the b asic cat io ns , p arti cularl y Ca , are lowered. Several factorial exp e ri men ts

22 HAWAII AGRICULTURAL EXPERIMENT STATION

were con d uc ted , using the very toxic Kilauea soil with vari ous fer tili zerco mb inati o ns as well as vari able amounts of several supplements. One su chexp erime n t involved 5 supplements: A-cal cium ca rb on a te , F- ca lciu m metasili cat e,G - silicic acid , O -sodium silicate, an d V-calcium sulfat e, at the rates o f 0, 3,6,and 9 to ns per acre of eq u ivale n t calc iu m carbo nate . Na2 Si0 3 , because o f its strongalkal in e nature, was ap p lied at one-f ifth the ca lciu m carbo n a te equ iva len t , an d this

was found to be too little.Fertilizer comb inations, all prepared to give 500 pounds per ac re o f N, P20S '

an d K20, wer e as follows :I (NI-I4hS04 (SA) , (NH4hHP04 (DAP),

II SA, CaI-IP04 (CaP)III NI-I40I-I (AA) , CaPIV Ur ea , CaPV KN0 3 , tripolyphosphate (TPP), urea

VI Ca(N0 3 h, CaP

All fertilize r com bi na t ions included KCI excep t no. V, an d all included MgS04at the rate of 100 pounds Mg/ acre.

Thus, this was a 6 X 5 X 4 experime n t with two replications for a to tal of 24 0po ts. Air-dried soil weighing 470 g was weigh ed into pl astic co n taine rs; thesup p leme n ts were adde d and t horo ughly m ixed into the soil; the fertilizers wer enext ap p lied either by pip ette or as solids ; th e whole mass was then water ed withdeionized distilled water , covered, an d allowed to in cubat e for 2 weeks a tgreenhouse temper atures-66 to 100 F. The co vers were t he n removed , th e con te n tsallowe d to dry and again thorough ly mixed an d finally planted to sudangrass. Eachcu ltu re wa s then placed in its randomly determined position. Afte r ger m ina tion, theplants were thinned to 5 per cu lture and watered as need ed with deionized distilledwat er. The plants were har vested on De cember 17, 1969 (Table 14).

Calcium metasili cat e is ou tstan ding in its effect o n toxicity e lim ina tio n ; sodiumsilicate is next, an d it is very likely that the yi eld m ight have been better had morebe en ap plied . The sodium sili cate ratoons wer e failures, and the so il pH, lik e that ofthe ch ec ks , indicated that the supp leme n t had been ex ha usted . Silicic acid had anap pare n t negative effect on the growth of the grass , having a lower y ield than theuntre ated chec ks (1.77 vs. 4 .0 5) . Ure a an d calciu m phosphat e were the bestfer t ilize r co m bina t ion; tr ipolyphosphate, and di ammonium phosphate mixes, werepoor. It is not possible to exp lain the very poor performance of Ca(N0 3 handcalciu m phosphate. The pl ants were com p le te failures and see med to suffer from anext re me phosphate deficien cy. In the ratoo n which followed , th ere was a co mp letereversal, and the no. VI fertilizer gav e the highest yield. Finally, the largest amountsof the supplement gav e the high est yields.

After the pl ant cro p was harvested, and be cause it was not possible to remix thesoi l without d estroying the roots, an application of ure a was m ade to eac h of thecultures and a ratoon crop was grown. It was soon apparent that the cu lture s whichhad grown well in the pl ant crop were growing less well, pointing to nutrientdeficien cies. The Ca(N0 3)2, CaHP04 fertilizer com bina tion gave outstandinggrowth , probably because it had overcome its earlier difficulties, whatever they

Su

pp

lem

ent

Tab

le14

.D

ryw

eig

ht

yie

lds

of

sud

angr

ass

(gra

ms

per

po

t)

Fer

tili

zers

Am

ou

nt

of

sup

ple

men

t(t

on

sp

erac

re)

~ C u: t'l

rJl o "l ~ > "l

"l

;:0 t'l o ~ t'" t'l r­ t'l > "l

IVU

rea,

CaP

9.1

3(c

)C

alc

ium

silic

ate

8.8

4(c

)II

SA

,C

aP8.

03

(c)

So

diu

msi

licat

e5.

98(b

)II

IA

A,

CaP

7.56

(b)

(c)

96

.66

(b)

Cal

ciu

msu

lfat

e5

.54

(b)

VK

N0

3,T

PP,

ure

a4.

71(a

)(b

)6

5.61

(a)

(b)

Cal

cium

carb

ona

te3.

92

(a)

(b)

ISA

,DA

P4

.06

(a)

34

.52

(a)

(b)

Silic

icac

id1.

77(a

)V

IC

a(N

03

h,

CaP

2.9

8(a

)0

4.05

(a)

HSD

2.61

HSD

3.24

HSD

2.17

o ... 5 ;:0 o tr: C;; > Z t) o ;:0 o ~ ;; t ;:0 t'l ""ce

24 HAWAII AGRICULTURAL EXPERIMENT STATION

were, an d the ratoon fertilizat ion in ad d it io n to the unused pl ant c ro p fertilizergave it advan tages over a ll the others .

The roots of the ratoon cro p were shake n as fr ee of so il as possible an d greenweights were taken . The root weights and the Ca a n d Si levels o f the tops are shownin T able 15 .

Ta ble 15. Green weight yields of ratoon suda ngrass roots (gram s per pot)with Si and Ca co mpos iti on of the tops

Percent dr y weighto f ratoon tops

Supplem ent

Calci um silica teCalcium carbo na teCalcium sulfateSilicic acidChecks

Average weighto f roots

4.0 9 (c)3.54 (c)2.96 (b)1.16 (a).90 (a)

Si

1.55.38.30.78.24

Ca

2.061.701.5 3.60.71

O f ve ry co n siderab le interest a re th e Si02 level s o f the ratoon tops.

Un fo rtu na tely, thes e w ere not determin ed in the pl ant c ro p, but it is quite obviousth at , w hi le the pl ants were a b le t o absorb Si from silicic acid , there was n o benefitto th em as reflected in grow t h either o f tops or o f roo ts . Th is d emonstrates rather

co nclus ive ly tha t Si02 • H 2°per se is not important in correc t ing the to x icity of the

so il so lutio n , but rather S i0 3 in co m b ina ti o n w ith some b a sic catio n .In su m mary , fr e ckling found o n sugarcane pl ants gro w ing o n ve ry acid , co ld , an d

p oorly d r ained Aluminous Fe rrugin o u s, llumic, an d I1ydrol Humic Latosolsre sults in m ajor p art fro m t he presence in t he soil so lu ti o n of fe rrous iron salts,

a lo ng wi t h sa lt s o f a lu m in u m , m anganes e, an d zinc as w ell a s so m e o ther m etal s .

Ro ot gro w t h is strikingly reduced, resulting directly in reduced t op gro w t h . One

may hypo t he size th at the abso rp t io n o f these e le men ts , p arti cularly ferro u s iron ,results in their accu m u la t ion in lo cali zed areas of the leave s w hi c h , in the a bse nce o f

adeq uate level s of ca lc iu m si lica te , be come n ecro ti c, particu la rly those leave s

ex posed to di re ct su n ligh t. Photo synthet ic act iv ity is t hereby redu ced,

compo un d ing the pl ant 's difficu lt ies.

Finall y , a wa rn ing n eed s to b e m ade . Where t he calc iu m m et asi licate is pi led in afield p ri or t o spreading, excess ive a m o u n ts m ay t hus b e app lied at that spot, and

th is wi ll re su lt esse n t ia lly in a growth fai lure of t he ca ne c ro p to foll ow . T he p lan ts

beco m e ve ry ch loro t ic an d so me sho w typica l magn es iu m- defic ie ncy sy m p to m s.

Affe c ted so il from su ch an are a was ta ke n for a p o t st udy . Ur ea w as u sed as a

so u rce of n itrogen for a ll cu lt u re s , and KCI w as added to giv e 200 po u n d s Nan d

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE 25

300 pounds K20 . MgS04 was ad ded a t 2 rat es : 1000 pounds per acre and 100pounds per acre . Other cu lt ures receiv ed one of th e following : ZnS04 , MnS04,CUS04 , FeS04 , or a co m b ina t io n of these. Wh ere used alo ne , each salt wasapp lied at the rat e o f 100 pounds per acre. In co m b ina t io n they were ap p lie d a t 20pounds per ac re. The results showe d very outstanding gro w t h for MgS04, but mostof t he others were co m p le te growt h failures .

Identifying Areas in Need of Calcium MetasilicateSeveral cr it eria are ava ilab le to th e growe r in identifying are as likely to resp ond

to ca lciu m metasili cat e.Firs t : The ap pearance of fr eckling o n ca n e lea ves growing in po orly d rained

aci d cl ay so ils, following co ld , w et p eriods, is a d efinite indication of need forCaS i0 3 .

Seco nd : Soil pH ran ge fr om below 4 .0 up to abo u t 5 .6 . A ny su ch so il, made upof iron and alumi nu m minerals , an d poo rl y drained, is ver y lik el y to respond. If t hesoi l is well dr ain ed but ver y ac id , the p lan ts mi ght very well sho w poor gro w t h dueto low Ca . If the Ca inde x is bel ow the crit ica l level, a cro p wou ld resp ond toCaS i0 3 , bu t wo u ld resp ond eq ually we ll to CaC0 3 , w hich is m uch cheape r. Eve n inpoorly drained are as showing fr ec k ling, w here the pH is suc h as to require a large

amou n t o f CaSi03 to bring the pH up to 5 .8 , for e xam p le , 15 to 20 tons, a progr amov er a p eriod of years o f ap p lying a nnual o r biennial incre m e n ts of I Y2 t ons o fCaS i0 3 and 1Y2 tons of th e much cheap er CaC03 will, in time , re m edy the problema t a reasonable annual cost. During later y ear s, o n ly the ca rb o nat e wo uld beneed ed .

Th ird : Plan t and soil a na ly sis o n aci d cl ay soils . T he positi ve id entifica ti on ofFe++ is a good star t ing point. Potassiu m ferr icyanide o r o r th op he nan t h ro line isusefu l here . A hi gh Mn/Si02 ratio in the pl an t tis su es also warns o f to x ic situat io n s.Ei ther so il or p la n t level s of Si are lik el y to b e only p ar t ia lly rel iable a t best. Itsee ms qui te cl ear that n o Si in e it her pl ant or so il is n eeded if there are no t o xi cage n ts pres en t in t he so il so lu t io n . If th e S i0 2 in t he so il is silic ic aci d , t hen eve nhigh level s woul d be of no valu e . Fa ilure t o f ind frecklin g o n Lo w lIumi c Lato sol sca nno t be cr ed i te d to th e presence o f S i0 2 in the irriga t ion water. Ra ther, undersuch soil co n d it io ns , Fe++ wou ld n o t accum u la te , h ence fr eck ling wou ld no t occur.

Fo ur th : Fin a lly , t he gro we r who sus pects so il to x icities can easily and qu ic kl yver ify h is sus p icio ns by a sm all po t expe r im en t , using bo t h CaC0 3 and CaS i0 3 asso il sup p leme n ts an d the fer t ilizer co m bina tion he o rd inarily uses.

MACADAMIA LEAF CHLOROSIS AT KEAAU )In co n t ras t to th e ve ry high ly weat hered so il at Kilauea , the subs t ra te at Keaau is

essen ti ally pu re lava . T he area chosen , ac t ua lly an o ld aa lava flow , w as co veredw it h a rel at ive ly luxurian t j ungle forest. In I 949, t he vegetation was knoc ke d down ,either bur ied or burned , the aa level ed , and young graft ed nu t t rees p lan ted in holes

I T he da ta and co nclusio ns in this section were fir st repo rte d t o th e ma na geme nt o f th eRo yal Hawaii an Macadami a Nu t Co m pa ny, Sep te m ber 19 68 . in oral and wr itten fo rms.

26 HAWAII AG RICU LTU RA L EXP ERIM ENT STA TION

o f abo u t 1.5 c ub ic fee t. S ince ther e was no so il as it is co m mo n ly known , th e yo u ngtrees were se t into so il taken from nearby farmland . S ubseq uen t field m an agementwas a form o f ro ck cu lt ure .

The su bs t ra te is m ade up of a thick lay er (as much as 15 feet) o f aa lav aov erlying pahoehoe lava and mi xed with t he accumulated organ ic debris resultingfro m ages o f decomposition of biological remains . The aa lava its elf co ns ists o fpi eces ran ging in size from rel ativel y fin e gra ins to large c linker like boulders. Theca t io n exchange ca pacity measu red 32.28 meq/l 00 g o f so il, an d th e total b asesat urat io n was 30 .9 p er cent , made up o f 17 .6 percent Ca , 10 .3 percent Mg, 2.2pe rcent K , and 0. 8 p er cent Na . Exchange aci d ity w as 22.27 m eq /100 g of soi l. T hearea receives more than 100 in ch es of rai nfa ll each year. Except in a fe w pl aceswhere p ah o eh oe lava is nea r the sur fa ce , th e natural drainage is goo d. However,becau se of management's desire to m echanize ha rvesting , th e su rface has beenvario u sly treate d an d co m p acted . It is thus not unlik el y th at anaerobio sis obtains ince r tai n ar eas for at least a part of the t ime .

A lea f ch lorosis began to develop on the o lde st trees some 10 to 15 y ear s aft erp lan t ing , appear ing first as a yello w ing o f the leaves an d th en th eir p rematureexcision . A fter a while so me of the m ost severe ly affe cte d t rees di ed .

Work was begun o n th e specific p roblem in A pril 1968 . T he orchardmanagement se n t sa m p les o f so il we igh in g 29 9 pounds w hich had been passedth rough a \4-inc h screen . To p rovide t hese sa m p les, 169 4 po unds of m at e r ial wasco llected fro m under selected o rc hard t rees an d fro m virg in forest nearb y . Of thistotal, o n ly 17 .7 p ercent would p ass th rough th e scree n, 82 .3 p erce nt was vario ussizes of rocks which were di scarded.

T he st udies we re made on the 17. 7 pe rcent fraction o f soi ls taken fro m under 12ch loro t ic trees, 2 no rmal t rees, and fro m the o r iginal virgin fo res t nearby . A t thestar t , t he soil was sampled at fo u r points aro und each t ree. These were co m bine dinto a single-t ree co mposite whe n it became o bvious th at no di ff erences ex iste d .For the large fac tor ia l experime n ts, t he en t ire lo t fro m u nder a ffec ted t rees w asco m po sit ed .

Foliar Analy sis ComparisonsT he leaf sa mp les sent from Keaau represented fo ur inte nsi ti es of ch lorosis, fro m

norma l to severe . Da t a from fer ti lize r experimen ts invo lv ing heal thy bearing t reeswere obtain ed from B. J . Cooil as part of experiments in st all ed a t Honomalino an dat Captain Cook , South Kona, Hawaii . Da ta we re a lso o b tained from no rm al t reesgro wing in Ka'u , Hawaii , from Masao Nakam ura, manager o f Haw aii an Orc ha rds,In c . These da ta a re shown in Tab le 16 . T he maj or elements , N, P , K, and S, a ll showa ri se in levels as ch lo ro sis intensifies . Th is is a typica l result where growth

restriction obtains, and usuall y indicates th a t no n e of th e accumulat ing nutrients isthe cause of the mal ady.

Cu and Fe decreased as ch loros is in ten sifi ed , sugges t ing possib le invo lve m en t. Mnwa s very low as co mp ared wit h t he o t her o rc ha rds. The Ca level for the normalleaves wa s h igher than fo r the o the r Keaau samples: .847, .46 7, .5 27, and .58 2 . Thesm all r ise of Ca in the more seve re stages was sometimes noted afte r seve re stu n t ing

o :> c en tTl

en o "'1 r­ tTl

:> "'1 "'1 ::e tTl o ;.:: r­ .trl r- ~ "'1 n ::;: t"' o ::e o en .En :> z "C) ::e o ~ .... "'1 > C: ::e tTl "" "

28 HAWAII AG RICU LTU RA L EX PERIMENT STATI ON

occurre d , i.e ., eve n the ele me n t in deficiency tended to accu m u la te withou tad d it io nal grow t h. Fe a lso wa s low in the leav es an d p ar all el ed ch lorosis, eve nthough the Fe in the so il so lu t io n w as very high. Zn at Keaau was ver y high an dbe lieved well in a to xi ci ty ran ge wher e Ca was not adequate. Mo w as two t o threet im es high er at Keaau.

From these data, it was possibl e to hy po th esize t ha t , wit h in the leav es a t least,Ca, M n, and Fe and possib ly Cu were in less than no rmal supp ly an d that Zn mi gh tbe in a toxic range. A t this point, it mi ght also be inferr ed t ha t the m ajo r p ro bl emwas a root prob le m because of to xi c soi l co n d it io ns , an d that fo liar defi cien ciessim p ly reflect t he in ability of the roots to p erfo rm m e tab ol ic ab so rp t io nadeq ua te ly .

Soil StudiesSoil extrac ts were made by taking 100 g (ov e n-dry eq u ivale n t ) of soil a nd ad d ing

20 0 ml of water. The virgin so il, because o f its flu ff in ess, re q u ire d 300 ml of water.T he flasks we re shaken occasio nally over a 48 -ho ur perio d , after which th ey wereautoclaved at 15 psi for I hour. Whil e st ill wa rm , th e sa mples were cen tr ifuged at3,700 RC F fo r 30 minutes. T he supernatan ts we re not co mple te ly clear, so th e ywere spu n again, but a t 32 ,000 RC F for 30 minu tes. T he ex t rac ts now were clearand were used for pH de te rm ina tions , bioassay, and ele me n ta l analys is on t h eatom ic absorptio n spec trop hotometer.

T he p ll of th e so il extracts from under affec te d tr ees Crab le 17 ) varied fro m4.25 t o 4 .95 . T he virgi n soi l ex tract showed 5 .60 a nd tha t o f the soi l from t henormal trees 5 .92 . A mar ked increase in soi l aci di ty is clearly associated wi th t hep ro bl em.

Bioassay co n sisted of germinat ing 20 w heat see ds (Genesee vari e ty) in 10 ml oft he extract in petri d ishes. T h e dishes were p laced in a warm roo m (92 F) and, after72 ho u rs with dai ly aeration, measurements of the longest primary root and t helo ngest shoot were made. T hese data are su m m ar ized in Table 17. A lt ho ugh t heresu lt s a re vari able , a ll of th e ex t rac ts showed some depression of growth . The poorgrowth in t he virgin ex t rac t su ggests that th e to xi c situa t io n already e x iste d at thest art of the orch ard as a result of natural we athering. Very marked depres sion ofgro w t h , as in the cases of 5 , 8, an d 15, sho we d stunting very like that cau sed w henp ure Fe+ + so lu t io n s were used. To check this point, d rops of ex t rac t were put ontoa spot pl at e, evapo ra te d to dryn ess a t moderate temperature , an d test ed wit hpotassium ferri cy anide so lution . In T able 17, heavy dark-b lu e co lo rat io n ,

cha rac ter is t ic of ferrous-ferr icy anide, w as shown ranging fro m zero to an intenseco lo r represen ted by ++++ .

When 10 9 of soi l were so aked in 3 0 m l o f Na 3 EDTA for 17 ho urs , th e co lo r ofth e resu lting ex t ract varied from a red brown t o ye llo w. Es t imates of the a m o unt ofFe in the e x t rac t (using orthophenanthrol in e) ranged from :~ 8 to 114 ppm. Theseresu lts also are shown in T able 17 and are high. They co m pare with 10 5 ppm forthe very toxic Ki lau ea so il (used) an d with 45 ppm for a moderat e ly tox ic so il atPaauhau (1'037) , both o f whi ch benefited from th e ap p licat ion of ca lc ium

met asilicat e .

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWN FAILURE 29

Table 17. Bioassay with wheat and other data (soil extract )

Soil Degree of Longest root X K ferricyan ide ppm of Fe in Na3numb er chloros is longest shoo t pH test EDTA extract

very severe 1 I.7 4.62 +++ 742 very severe 22.0 4.50 + 443 very severe 15.2 4.58 ++ 614 very severe 17.5 4.43 +++ 825 very severe 10.5 4.32 ++++ 1146 very severe 19. 6 4.40 +++ 747 very severe 19.3 4.84 +++ 718 very severe 8.9 4.35 ++++ 1039 severe 12.2 4.56 +++ 92

10 very severe 16.2 4. 55 ++ 3811 sev er e 12.5 4.2 5 +++ 11412 very se vere 15.8 4.80 ++ 5213 none 14. 3 4.95 ++++14 none 22.0 5.92 0

15 virgin soil 9.3 5.60 +16 disti lled H2O 27.3

To tal An alysis of Aqu eou s ExtractT he so il ex t racts we r e nex t analyzed for the var io us e leme n ts shown in Tab le 18,

an d are reported as ppm of t he d ry so il but refer on ly to t he water-so lub le p ort ion.Ca and Mg a re ve ry va r iab le and suggest a t least so me deficie nc ies . T here is anappare n t negat ive co rrelat io n b etw een the w heat -seedling growt h and th e Fecon te n t of the extract (- .628*) and also wi th the A l co ntent (-.466 1 ° ). Z n in theo rc hard so il is from abou t two to five t im es h igh er th an in t he virgin soi l and ca nalso be a part o f t he t o x icit y . In separa te st ud ies, the 50 percent po int in growthdepression of wheat seedl ings o ccurred at 10 pp m for Fe++ and 15 to 19 pp m forA I, 15 for Cu, 5 for Ni, 30 for Co, and 15 for Zn .

T he soi l from under tree no . 13, whi ch showed no ch lorosis as yet, m arkedl yretarded wheat growth , and this fa ct together with t he abundant Fe++, as refl ect edby th e fe rri cyanide t est and by t he markedly h igher acidity , p ro bably indicat es tha tch loro sis would de ve lo p soo n if these co ndi t io ns were not co rrec ted. The highlych loro t ic soi l no. 2 , whic h gave good wh ea t growt h, showed a low ferricy anide testand lo w Fe++ analysis , but also sh o w ed a very low Ca level. T h is also app lies to no .10 and 12. Soils no. 1, 5, 8, and 12, wh ich gave t he poorest growth with w heat andwhi c h produced very severe chlorosis wi th m acadami a , ga ve th e h igh est Fe ++re adings (Tables 17 and 18 ) even th ough t he fi rs t three had hig h Ca read ings . T hus,these data ind ica t e t hat t he problem is assoc ia ted wi t h hi gh Fe++ and /or lo w Ca .However, even though Ca may b e adequate if very h igh levels of Fe++ exist,toxicity (m a inly to the roots) can be le thal.

30 HAWAII AGRICULTURAL EXPERIMENT STATION

Ta ble 18. Wat er-solub le elements of the dry soil (ppm)

Soi lnumbera Fe Cu Mn Ca Mg AI Si Zn

1 45 .0 .4 10.4 70.0 35 .0 8.0 110.8 2.22 11.4 .4 2.9 27.0 13. 1 3.4 86 .2 3.73 21.8 .5 3.5 55.4 33. 1 9.8 87. 2 2.04 27.8 .3 10.6 54. 2 30.8 5.4 102.0 2.05 46 .8 .5 8.4 109.4 66 .8 12.0 128.0 2.66 32.0 .3 9.6 57.4 38.4 3.6 106.8 2.47 16.0 .6 2.6 42.4 29.9 13.8 76.8 2.58 49.6 .5 10 .4 108.4 54 .7 10.8 135. 6 2.49 29.6 .4 4.0 121.2 96.6 22.4 162 .2 3.4

10 8.0 .6 2.4 37.4 13.5 4.6 70 .2 1.011 37.2 .3 6.0 103.8 75.4 22.4 171. 2 2.812 9.8 .4 1.8 35.8 28 .0 9.6 93.4 1.013 32.4 .4 9.8 67. 2 67 .6 9.4 137.4 3.014 1.6 0 1.2 45.8 7.9 2.0 132.4 0.715 12.0 .2 5. 1 149.0 68 .4 5.0 87.6 0.6

aSee Table 17 for severity of chlo ros is.

I t se ems rathe r clear so far t hat, in t he opera tion ot t h e orc hard over t he 15 -y ear

period, t he pra ctices fo llowed tended t o increa se ra t h er substantia lly t h e soil ac id it yfrom 5 .6 to aro u nd 4 .2 to 4.6 and caused the in cr ease in t h e am o u n ts of Fe an d A l

in the soil so lu t io n whi le reducing the Ca an d Mg lev els . Perhap s also b e cause of

efforts of man agement to d ev e lop a hard an d sm o o t h so il surface to faci lit ate

ma chine harvest, some tendency toward lim it ing a ir penetration into the so il

re sulted in at least partial anaero b ic con d it io n s , permitting greater accumulation of

th e very toxic ferro us iron. From the an alys is of the ex t rac t of the virgin soi l, it is

cl ear t hat ncar -to x ic levels (for wheat) a lready ex is ted a t th e start.

A review o f fer til izer p ractices fro m t h e star t of the orchard showed t ha t

su lfate of ammonia (S A) was u sed from 1949 t o abou t 196 0 . During this period

vario us minor cl ements a n d magn es iu m were ap p lie d t oge t he r wi th su pe rp ho sp ha te(CaP) . Fro m 19 6 0 , the SA was reduced in quantity bu t all t h e ca lc iu m phosphate swere rep laced with Dimon (DAP) .

In retro spe ct it se e ms cl ear that the co n t in ue d us e of SA, followed by th e shiftaway fro m calc iu m pho sphat e to DAP , su bjec te d the tr ees to large a mounts o fresidual ac id ity . Fina lly , the e lim ina t io n of a ll Ca would h a st en the accu m u lat io n ofto xi c ele me nts w h ile re d ucing t he abi lity o f the nutrien t-absorp tion m echan ism toprevent th e upt a ke o f the to xi cants . 11 is a lso possible that t he con tinued u se of at

leas t so me minor clements re su lted in so me excesses.

CAUSES OF LEAF FRECKLE, LEAF CHLO ROS IS, AN D GROWTH FAILURE 3 1

Soil Culture StudiesIn ord er to ch ec k o ut the various po ssibi lit ies developed so far, an d af te r some

preliminary attemp ts, a factori al experiment was d esigned. Six fertili zer m ateria lswere us ed, each one con tain ing K2S04 (except no. VI) as well as MgS04 : I-SA,DAP ; II-urea, DAP; III -SA, CaP; IV -urea, CaP; V-Ca(N03hm CaP; VI -potas­siu m nitrate (KN03 ) , urea, CaP.

All the fertilizer mi xtures were m ad e of C.P . reagent chem ica ls an d werecalcu la t ed to give 500 pounds N, 17 5 pounds P, 29 0 pounds K , an d 10 pounds ofMg pe r acre.

T hree supple me n ts or soil co rrec t ives from C.P. che m icals we re app lied a t fourlevels eac h : A-calci u m ca rbonate, B- calcium m etasilicate, and C-magn esiumcarbo nate . The amo un ts of eac h used were calcu la te d to give the neutrali zing powero f 0, 5 ,10, and 15 tons o f CaC03 per ac re.

The minor e lemen ts , Mn , Cu , B, Mo, an d Zn, were also included an d werecalcu lated to give 5 pounds o f element per ac re and Mg at 10 pounds per ac re .

The experime n t wa s designed for six fer tili zer mixtures, three sup p leme n ts atfour levels an d on e minor ele men t treatment with two replications , an d thusbecame a 6 X 3 X (4 + I) X 2 factorial fo r a tota l of 180 pl ots. Each pl ot was aI -pint pl asti c co n tainer cap ab le o f holding 55 0 g o f o ve n-d ried equ iva le nt Keaauso il. A single 3-m m hol e abo u t halfway up on one side pro vid ed ad equat e drainage .

So lub le fe r t ilize rs an d minor e leme nts wer e m ad e up in stock so lu tions an dapp lied by p ipette . The supp leme n ts an d the calciu m phosphat e wer e weigh ed o u tto 0 .0 I g and app lied as dr y m at erial. In a ll cases, afte r the fe r t ilize r an dsup p le me nts were app lied , the so il was thoroughly mi xed an d returned to thecon tainers , wat ered with distilled wat er, seal covere d , an d allo we d to in cubate a tgreenhouse temp er atures (70 to 100 F) fo r 2 weeks, afte r which the co n tainerswere uncov er ed an d allowe d to dry out to a moder at e degree. The so il again w asthoroughly mixed, returned t o the pot, see de d with fiv e cucum be r see d s, an dpl aced in the green ho use in a comp le te ly ran do m ize d m ann er within ea ch block.Watering as needed was d on e with di stilled wat er. The see d lings wer e thinned tothree per pot.

Twen ty-five days la t er the ex p eri me n t was har vest ed b y cutting the pl ants justabove the soil lin e. The green weights were obtained at o nce . Results are shown inTables 19 , 20, an d 2 1 as ave rages per pot. In each su m m ary the re p lica t ionsrepresented p er it em may be obtained b y dividing 180 by the number of items.Thus, within T able 19 ther e are 18 it ems, m eaning that each o n e is an aver age of 10separa te p lots . The three aver age s underneath represent 60 each, and the sixaverages in the vertical co lu mn at the right represent 30, et c .

Varian ce analys is revealed highly significant differ ences am ong fertilizers, an dam ou n ts o f supplement. Differ ences am ong the supp le m en ts wer e not sign ific an t.Minor ele me n ts reduced gro w t h for ferti lizer s I to V , but gave the highest ac t ual potyie ld with fer t ilize r VI.

T he su mmary of th e fertilizer re sponses is show n in Table 22 . Yields forfertil izer tr eatments varie d from 22 .6 for fertilizer III to 44.0 7 gra ms fo r VI.Fertilizer s V and VI, mad e up o f the basic ca tions Ca, K, and Mg, but co n ta ining noam moni u m and on ly a sma ll amo u n t o f ur ea in one case , gave the highest y ields . It

32 HAWAII AGRICULTURAL EXP ERIMENT ST ATION

Ta b le 19. Gree n weights (gram s per po t) for fertili zer X supp leme nts

Sup pleme nts

CaC03 CaS i03 MgC0 3

Ferti lizers A B C Average

I SA,DAP 25.1 26 .9 24.7 25 .5II Urea,DAP 36.5 37 .8 28. 1 34 .1

III SA , CaP 23 .6 21.3 22.9 22.6IV Urea, CaP 38.4 39 .0 42 .3 39.9V Ca(N03h, CaP 40.6 43.4 47 .9 44 .0

VI KN03, ur ea, CaP 49 .0 42 .6 40 .6 44.1

Average 35 .5 35.2 34.4

Tab le 20. Green we ights (grams per pot) for sup pleme nt X amounts and minor elements

Tons of sup p lement per ac re

MinorSupplement element 0 5 10 15 Average

A CaC03 25.5 32.7 42 .0 37.7 39 .8 35 .5B CaS i03 30.4 30.8 37.0 37.9 39 .9 35.2C MgC0 3 27. 1 32. 9 34.6 40 . 1 37.4 34 .4

Average 27. 7 32 . 1 37 .9 38.6 39. 0

Ta b le 21. Gree n we ight (gram s per pot) for fer t ilizer X amountof supplement and minor clement

Tons of su pplement per acre

Min orFertil izers elemen t 0 5 10 15 Average

I SA, DAP 24 .8 14.0 31.3 34 .7 23 .0 25.6II Urea,DAP 29.5 17.7 32 .7 41. 3 49 .5 34.1

III SA, CaP 13.8 2 1.5 22 .7 28.2 26 .8 22.6IV Urea, CaP 35 .7 34 .5 42.8 41.5 45 .0 39.9V Ca(N03h, CaP 38.5 40 .3 49 .8 44 .8 46.3 44 .0

VI KN0 3, urea, CaP 50.2 38.0 47.8 40 .8 43.5 44.1

Average 27.7 32. 7 37.9 38 .6 39. 0

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE

Table 22. Gree n weights (gram s per pot) for fertilizers X amo untsof sup pleme nts and min or ele ments

33

Trea tment

IIII

II

IVV

VI

Fert ilizer combination (all inclu dedequal amo unts of MgS04 and K)

SA, CaPSA, DAPUrea, DAPUrea, CaPCa(N03l2. CaPKN0 3, urea, CaP

HSD

Tabl e 23. Gr een weight (gram s per pot) for sup plem ents

Green weight yie ld(grams per pot )

22.6 (e)25 .5 (d)34 .1 (c)39.9 (b)44.0 (a)44.1 (a)

2.7

Suppleme nt

A

B

C

Calc ium carbona teCalcium silicateMagnesi um carbo na te

Green weight yield(gram s per pot)

35 .535 . 134 .4n.s,

is cl ear that the y ie lds de cline as the amou nts of a m m o n iu m sa lts a re in creased and

as the Ca is reduced. F ertilizer III , with calciu m phosphate , w a s le ss goo d than

ferti lizer I , w ith diammonium phosphate , probably be cause a ll of the N in HI was

from the su lfa t e.Variance analys is sh ows no difference s in y ie ld am o ng the three supplements

(T able 23) .There w as, however, a highly sign ifican t resp o nse to amounts o f the su p p le me n ts

(T able 24) .Thus, the 5- t o n treatment of anyone of t he supplements ga ve a highly

significant re sponse over none, but th e minor e le me n t s definitel y depressed grow th ,

the lo ss being hi ghly significant (1" = 155.50**) . None of the p ossible in terac ti on s

ac h ie ved significance .

Effect of Treatments on Soi l pHA fter the crop wa s h arvest ed, soi l pH d eterminations w ere m ade ('rab le 25) .

Fer t i liz e rs had a very sign if ican t effec t o n so il pH (1" = 20.73**). Thus, the pH o f

the p articular co m pos ite d so il u sed had an initial pH o f 4. 95 . Fe r t ilizer I , whi ch h ad

34 HAWAII AGRICULT URAL EX PE RIME NT STATION

Tabl e. 24 . Response to amo u nts of suppleme nts

I

IIIIVII

VI

V

Amo unt of suppleme nt(t ons per acre)

o + minor clementso5

1015

HSD

Table 25 . Effec t of fer ti lizer s o n so il pH

Fert ilizer s

SA, DApSA, Cal'Urea, Cal'Urea, DAI'KN0 3, ure a, Cal'Ca (N0 312, Cal'

USD

Gree n weight yield(gram s per pot)

27.7 (c)32.7 (b)37.9 (a)38 .6 (a)39 .0 (a )

2 .3

Soi l pH

4.58 (c)4. 70 (b) (c)4. 78 (b)4.87 (b)4. 87 (b)5. 17 (a)

. 18

the N in the form of ammonium su lfa te and di ammonium phosphate , was themo s t acid . Ferti lize r Ill , whi ch had o n ly SA and Cal', was next. T hen fer ti lize rs III ,IV , and VI , in whi ch calcium phosphat e provided the pho sp hate, an d II, co n ta iningurea an d DAp , wer e gro u pe d fro m 4.70 to 4 .8 7. Ferti lizer V, wi th ca lc iu m n it ra te,ca lc ium pho sphat e , potassium su lfa te , an d m agn esium su lfa te, actually adde dbasicity to t he soi l, as would be ex pected, an d is superior to VI in which potassi u mnit rat e and ur ea provided the N.

As with yield , the three supplemen ts which were ap p lied acco rdi ng to a uniformneu traliz ing power had th e sam e effe c t on pI!. T hus , A-4.88, 13- 4 .8 2 , and C- 4 .8 2are st at ist ica lly the same .

T he amo un ts of supplements used , however , affec te d th e p H sign ifica n t ly , T ab le26 (1" = 20 .54 **) . It shou ld be remember ed t ha t on ly th e " fin es" of the so il wereused an d t ha t this co ns t it u te d 17.7 perce nt of the total. T he rela tively small changein the pH readings points to a very stable buffer sys te m in the so il fr action. Theminor element cu lture sho we d a pH of 4 .79 . Undo ub te d ly , the regression on pHwould have bee n steepe r had more t im e elap sed.

CAUSES OF LEAF FRECKLE, LEAF CHLO ROSIS, AND GROWTH FAILURE

Ta ble 26. Effect of amo u nts of supp leme nts on soil pH

35

Amo unt of supple me nt(t on s pe r ac re ) So il pH

0 4.64 (c)5 4. 75 (b) (c)

10 4.84 (b)15 5 . 12 (a)

--HSD 0.1 4

T he re was an in te rac tio n between fertili zer and am ounts of supplements (F =4 .91 **). T he data are sho wn in Tab le 2 7. Ver y clearl y, the SA · and DAP -fer tilizedp lots (I, II , and III) res ponded much mor e slowly to n eu trali zation t ha n did theo the rs. Fertilize r V , in which Ca( N0 3 h provide d all the nitrogen, had th e steepestreg ression.

To ga in so me fur t he r kno wledge of the buffering capacity of the so il, themat er ial whic h passed th e ':4-inc h screen (t his was 17.7 percent of the o rigina l) wasscreened again throug h a 4-m m screen . Most of the m ate rial (8 1.6 pe rce nt) passed,thus representing abo ut 14 pe rcent of the origi na l fie ld ma terial. Treating 25 gramsof so il ( 1:5) with qua ntities of Ca(OHh , ran gin g in 40-mg in cremen ts fro m zero to200 . gave t he d ata shown in Tab le 28 .

In ap p lying t hese data to fie ld app lications , con siderab le experimentation isne ed ed, for it must be emp has ize d t ha t the fin es used represen t o n ly about 14percent of t he total field mass. If 9 .35 tons we re ap p lie d per acre , a m uch greater

Table 27 . So il pH intera ct io ns for fer ti lizers X amounts of suppleme nts

To ns of suppleme nt

Fer ti lizer 0 5 10 15

I 4.59a 4.25 4.55 4.7 3II 4.66 4.89 4. 78 5.04

III 4.94 4.65 4.50 4.80IV 4.42 4.61 4.8 2 5.23V 4. 59 5.22 5.46 5.74

VI 4.70 4.85 4.93 5.2 1

aEaeh da tum is the average of six sepa rate determinati on s.

36 HAWAII AG RICU LTU RA L EX PERIMENT STATIO N

Table 28. Buffering capaci ty of the Keaau soi l fines

Ca(OHh(m g/ 25 g of soi l)

o4080

120160200

Equivalen t tons CaC03per ac re

o1.873.745.6 17.4 89.35

Soil pH(1- 5)

5.756.106.426.706.877.20

pH change co uld res ult, since so far as the fine frac tion is co ncerned th is w ould bethe equivalent o f so me 65 tons pe r acre, and this co u ld ca use m ajor prob lems.

General Conclusions and RecommendationsThe prob lem of ch lorosis at Keaau is t he d irect resul t of two main t hi ngs : t he

nature of the virgin substrate to begin with, an d the co ntin ue d use of am moniacalfert ili zer salts without ap propriate Ca addi tions to neu tr alize the accum u latingac id ity and to fortify the tree' s ability to withstand t he seve re toxicit ies o f Fe++ andA I. Seconda ry facto rs include the possibili ty that the continui ng effo rts at m akingthe surface smooth by ap p lyi ng volcanic cin ders and p ressing t he m in has developeda su rface seal restraining free access of air to the rhizosp he re, resulting inaccumu la tion of Fe++ in the so il so lu t io n and re duct ion of roo t vigor. Consideringthe mar ked setback in growth of the cuc u mber p lan ts by the ad di tion of minorele me n ts, it is likel y t ha t th e prolo nged use of the several m ino r elements added tothe total to xicity which fina lly ove rcame th e roots.

T he so lut ion to t he prob lem ra ther clearly requires a shift away fro mac id-prod uc ing fe rt ilize rs and toward a Ca regime. T his ca n be accomplished b yusing ei the r sing le or trebl e supe rp hos p hates and a j udicious b ut co n ti nui ngp ro gram of ca lci u m ca r bonate (coral stone ). Urea w ould y ield less res idual aci d itythan eithe r the sulfa te or phosphate of ammonia. So me effort shou ld beund er taken, ex perime ntally a t least , to roug h up the surface af te r each harvest tofac ilitate air movement in to the so il. Also, so me effort should be ma de to mi x thecarbonate as well as the sing le or tre b le su perp hosp hate into a t least 2 to 3 in ch es o fthe rocky me d iu m , and t h us enab le the surface roots to establish actual co n tactwith the Ca.

When t he Keaau pro b lem is co mpare d with the Kilauea frec kling, the similar it iesin soil so lute co mposition cause one to wo nder why each of t he three supplementsused was equa lly effec tive, while at Kilauea, only t he silicate was effective. Apossib le hy pothesis wh ich canno t now be p ursued m ight we ll be associated with thedegr ee of weathe ri ng of t he t wo media, res u lti ng perha ps in the surfaces of t heKilauea soi l particles bei ng as to xi c as the sol u tion itse lf. T hus, t he Si0 2 con tent of

CAUSES OF LEAF FREC KLE, LEAF CHLOROS IS, AND GROWTH FAILURE 37

the aa lava be ing much higher than th at o f the highly weather ed so il suggests thatsurfaces in the one case are mo stly Si02 , whi le in the o t he r , Fe+ +, Fe+++ , an d A Ipr edominat e .

MAUl SUGARCANE GROWTH FAI LURE

Maui grow t h failure (MG F) is a name give n a malady of sugarcane grow ing ince n tra l Mau i. As the name suggest s, co mp lete gr owth failure o f so me varie t ies hasbe en observed. During the early 1920's, McGeorge (16, 17, 18, 19) recognized thatthere was an appa rent Ca deficien cy and undertook var io us methods o f ac id ify ingth e otherwi se mi ld ly alka line soi l, but without success.

Martin (14), in hi s book on suga rcane di seases, wrote in part : " T he earlysy mp to ms [o f Maui Growth Fai lure) ar e recognized by a slowin g up o f the growthof the cane an d by the presen ce in the leaves of o ne or m ore wid e ch lo ro tic str ipe sthat often extend the length o f the leaf .. .." He went on to observe t hat somevari eties withstood th e malady better than others , but sugg ested n o ca usal orremedial factors.

In t he 19 51 R eports of the Hawaiian Sugar T echnologists , Cleme n ts (1) wrote ,"T he so -ca lled Maui Growth Fai lure at Hawaiian Commer cial an d Sugar Co m pa ny isanot he r [unexplained) phenomenon .. . . Here , varie ty H 37-19 3 3 is a failure whileH44-3098 do es pretty well , and 11 38-291 5 is intermediat e . S heat h analyses forsome va riet ies growing in the MGF area are:

Varie ty K index Ca index Mg index11 37 -19 3 3 3.40 .138 .49811 38-291 5 2.72 .0 77 .343H4 4-3098 3.26 .048 .23 1

" . . .I t see ms lik el y tha t growt h fai lure, of 19 33 parti cularl y, is due tomagn esium toxici ty, and tha t 3098 do es better because it absorbs less magn esiumth an do es 19 33 . T he ca lci u m levels are co nsi de ra bly below normal."

So far the o n ly so lu tion to this problem has be en th e se lec t ion o f t ol er antvar ieties wh ich give sa tisfac to ry yie lds a lth oug h lower than they do in unaffe ctedareas .

Tissue Analy sisIn 1968 a sea rc h was begun to identify the problem an d hopefully find a

so lution . A half ton of soil was air-d ried, screen ed, an d thoroughly mix ed by theplant at ion (Hawaiian Commer cial an d Sugar Company) and sh ipped to theUniver sity of Hawaii, Department o f Plant Ph ysi ology . The soi l is a Low HumicLatosol made up of a silty loam, often co ntaining gravel an d cora l sand. The so il h asa ca t ion exc ha nge ca pacit y of 2 1. 12 meq/lOO g of whi ch 29 .3 per cent is Casa t ura te d, 39.4 per cent Mg, 8. 6 pe rce nt K, and 8 .2 per cent Na fo r a base sa turat ionof 85 .5 per cent.

2Esse n tially thi s same rep ort was present ed to the man agem ent of Hawai ian Co m me rcia lan d Sugar Co mpa ny on April 29 , 1969, in writ ten and oral forms.

Tab

le29

.C

rop

log

dat

afo

rn

orm

alpl

ants

and

pla

nts

dist

ress

edb

yM

aui

gro

wth

fail

ure

Pe r

cen

tP

erc

ent

Per

cen

tgr

een

dry

tiss

ueP

erce

nt

Part

sp

erm

illio

nw

eigh

tw

eigh

tm

oist

ure

suga

r-fr

eedr

yw

eig

ht

suga

r-fr

eedr

yw

eigh

t

--

-M

n/S

i02

Fie

ldV

ari

ety

Gro

wth

H2

OT

SL

NK

-H2

OA

PI

Ca

Mg

SSi0

2M

nZ

nC

uA

IM

oF

eB

rati

o

901

H50

-720

9go

od84

.510

.01.

83

.660

4,1

34.0

80.1

15.2

35

3.16

236

329

11.8

93

14.

27

819

H50

-720

9b

ad82

.212

.21.

56.7

02

13,5

00.0

59.1

35.2

231.

99

4159

1410

1.14

295

.821

300

H5

7-51

74go

od82

.39.

71.

49.5

306,

440

.254

.077

.129

2.05

109

3016

12.5

59

5.8

538

20

H57

-517

4b

ad76

.217

.41.

10.6

16

6,8

34

.08

8.0

94.1

591.

78

1933

1913

1.21

27

6.8

11

'"co :::t: > ~ > o :;.:l 2i c t; c :;.:l > r- M X ." M :;.:l §: M Z ..., C

/l ..., > ..., o z

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE 39

At the same time , the plantation co llec ted cro p log samp les of tw o fairl ytolerant vari e ties (H50 -72 09 and H57 -51 74) from MGF areas and also from go od

land whi ch is capab le of maximum yields for th e area and sent them to th e BrewerCrop Log Laboratory for co m plete analy sis. T he results in Tab le 29 show thedomin ant problem to be asso cia te d with the very low Ca re adings. T he leaf Nreadings in the bad areas are obviously ver y low and no doubt contribute greatly tothe poor growth . In th e area , in general , aqua am m onia has been the only fertilizerused, since both P and potash levels are very high. The very low readings in the badare as may well reflect a poor root system as well as a loss of this vol atile fo rm of Nfrom th e warm and somewhat alka line soil (pH 7.5 to 7.9). The very low tissuemoisture levels refl ect not only the poor root syst em but also the low N. The tissueP readi ngs are very high , as are all the K-H2 °readings. B is rather high. Mg is abou tnormal , although so mewha t higher in the poor are as . Zn is quite high , Cu is normal ,but Mo is higher in the poor areas. Mn and S t end to be low in both conditions, an dFe is quite low for both the good and bad growth area s, probably a re sult of thealka line so il. T he total sug ars of the sheath are high in the affected areas-areflection of th e low N, low moisture, and poor growth . The low Mn/Si02 ratios of7 and o f 11 may well sugges t need for Mn in both good and bad areas .

In examining these data, it must be app reciated that the two varieties reportedon , especially H50-7209, are fairly tolerant of th e m alad y. The sy m pto ms as seen inthe field no w are somewhat differ en t th an those reported above by Martin in 19 38.The yo unger leaves of an o ccasional plant are very markedly st riped , not unlike thatnoted for Mn deficiency , and also not unlike that seen on young plants in a seriousS defi ciency (5). The older leaves of some of the 6- to 8-month-old plants havech lorotic stripes through out th e length of th e leaf but sca ttere d irregul arl y ac rossthe leaf , leaving so me broad an d some narrow green bands in between. T hissymp to m also 'has been ob ta ined in m inus-sulfur sand cu lt ures as we ll as minus-zin ccu lt ures .

Soil ExtractsDistilled water and IN HN0 3 ex trac ts (1 :3) w ere made of the MG F soil.

Immediat e and co p io us fro thing fo llowing th e addi tio n o f acid indicat ed th epr esen ce o f a grea t d eal of ca rbonate-b ica rbonate m at er ial. Atom ic absorp tionscan n ing o f the ex trac ts was undertaken, and the co ncen tratio n of the observedelem en ts determined . T he results are shown in Tab le 30 .

Ta ble 30 . Eleme nta l composition o f 1:3 extracts (ppm)

Ex tra ctant K

1730 1

Na

742 17

Ca

3.718.3

Mg

3.87.3

Fe

0.348 .7

Zn

0. 12.8

Cr

o1.8

Ni

o3.3

Co

0.18.2

40 HAWAII AGRICULTURAL EXPERIMENT STATION

It ap pears quite obvious that the high Na an d K would have much to do with thelow tissue Ca whi ch, in turn , co u ld in crease th e Mg level noted earlier forsus ceptible varieties. Of interest is the high K-H 2° reading, shown in Table 29, forth e sh eath tissue. The Na index of th e same tissue is only 350 ppm , but co ncurren tstudies in sand culture show the Na to accumulate excessive ly in the roots andstems before rising eve n a little in the sheaths. Because of the levels of themonovalent cat io n s, the plant finds it difficult to absorb adequate Ca . It is alsopossible that some elemental deficiencies e x ist , particularly Mn , and perhaps a lso Sand Zn. Whether this is so because of a lack of the two elements in the soil, or aprevention of their absorption because of the ca t io n disbalance and its effect on theabsorption m echanism used by sugarcane, is a part of the problem.

LeachingBecause of the Na in the soil solution, the first obvious thing to do was to leach

the soil. A 28 -q uar t plastic basket with four small drainage holes in its bottom wasfill ed with the soil and leach ed with di stilled water for 3 months. The Na co n te n t ofth e parti cular lea chate was 116 ppm at the start, 74 ppm on August 20,53 ppm onSeptember 9, 41 ppm on September 16, and 14 ppm on O ctober 14.

Preliminary t ests with sud angrass sho wed a marked growth improvement on th elea ched as co n t raste d with th e unleached soil. I mprovement w as also obtained whenthe original soil was treated with 5 tons per acre of gypsum, ano t her co m m onmethod used in co r rec t ing high-sodium soils. It appeared also that a minor elementproblem d id in fa ct ex ist and th at the parti cular fertilizer combinations hadsomething to do with determining which minor ele m e n t in each case was deficient.Therefore, a rather complicated trial was set up, aimed at getting more Ca into th eplant and at growth improvement. The general plan of th e ex per ime n t calle d forthree so il prepar ations: (I ) the untreated so il as received from the pl antation, (II)that soil thoroughly mixed with very fin ely ground gy ps u m (5 tons per acr e) , and(lll) that soil after leaching with distilled water for 3 months with th e Na level in

th e leach ate finally down to 15 ppm.Sev en fert ili zer treatments were set up. In each ca se the application ra tes were

500 pounds of N, 218 pounds of P, a nd , excep t in ce r t ain ones, 5 2 pounds of K p er

ac re :A. (SA ), (DAp ), a nd KCI. T his mixture would leave an acid residue and might

thus enc o urage the loss o f sa lts and also pro vide a sta b le su pp ly o f Nan d S.

B. CaCN2' (CaP)' and KCI. A heavy su pp ly of Ca was thus provided.C. Ca(N03l2·41120 (Ca( N03l2), Cal', and KCI. Like 13, this would provide a

large amo un t o f Ca.D . Ure a , Cal', a nd KCI. Urea was included since it is a co m m o n com me rc ia l

fertiliz er us ed by sug ar pl ant ations.E. Ur ea, Cal', but no KCI. KCI was o m itte d because of th e high level of K

a lready present. It was included in th e o t he r m ix tures so th at eve n th e

leach ed so il would not be defici ent.F. Urea alo ne. Since the ar ea by both tissue and soil analys is is classed as a ver y

high -phosphate a rea as well as a high -potash area, this see me d a reasonabl efert ili ze r. I t would hav e been bett er t o include a N 114 Oil treatment, since it is

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND G ROWTH FAILURE 4 1

the fertili zer used by the pl antation. (This was included in a lat er t e st, seebel ow.)

G. CaCN 2, Ca P, and KN0 3 . T his is a moderately h igh-calcium fertilizer whichgives rise in the soi l to ur ea, and co n tains no chlo ride .

Fina lly , the third factor, t he minor elemen ts and Mg were set up as fo llo ws: (1)check , (2) ZnS04 , (3) CUS04, (4) MnS04 , (5) MgS04, (6) H 3 B0 3 , an d (7)Na2Mo04 . Applicati ons were made to provide 25 pounds p er ac re of th e eleme n t ,excep t for Mo an d thi s was ap p lied at 10 pounds per acre .

T hus, this ex p eriment , a 3 X 7 X 7 with two replications, ca lle d fo r 294 cult u res .T he pots us ed were I -pint sq uare pl asti c co n taine rs which h eld 780 g of theair-d ried soil. After all the che m icals were adde d, th e soil was water ed with di stilledwater an d the pl ast ic pots were seal-co vered and allowe d to in cubat e . Afte r 2 week sat gree n ho use temp er atures, the covers wer e rem ov ed and t he so il was allo we d t odry ou t , after whi ch ea ch cult ure was emptied o nto clean paper, t h e con te ntsthoroug hly mi xed again , then see ded wit h sudan grass, and placed in its randomlypred etermined position.

Through out t he gro w t h pe rio d, wa te ring wi th distill ed wat e r was done asneed ed. Germi na tion was excelle nt an d all po t s wer e thinned to five pl ants. Ingene ral, growth was good and no thing inter fer ed whi ch precluded ge tting cogentdata refl ecting the im p act of t he several t reatmen ts. The sudang rass w as harveste don Ap ril 14, 19 69. Green weigh ts were de termin ed a t once an d prepared foranalysis. A su m ma ry of the dry weigh t y ields in rela tion to treatment fo llows inTa ble 31.

So il treatment gave h ighl y sign if ican t yield d ifferences (F = 4 1. 19**) for eachtreatment with lea ch ing> gy psum> ch eck . Fertilizers also showe d highl y significan tdi ff eren ces (F = 188 .6 3**) . The ord er was as follows: A> C = D> E> G = B>F. Forthe minor elements (F = 8.44** ) the ord er was Zn>Cu = Che ck = Mo = Mn =Mg>B . T hus, in the overall , Zn was t he o n ly elem en t superio r to the check , and Bt he on ly eleme n t whic h depressed growt h , alt hough , as will be seen later , certaininterac tions ch an ge t his. It m ay , th erefore , be st at ed at this point that Mau i gro wthfai lure can be overco me by leaching out t he Na , using the proper fe rt ilizerco mbin ation, and applyi ng at least one m inor elemen t, Zn . It is also po ssible that B,because it depressed growth, may already be prese n t in ex cess ive am oun ts.

Soil t rea t ments and fer til izers showed a significant (F = 2.54**) in teraction ,alb eit rat her weak , Tab le 32 .

T hus, the A fer t ilizer (SA , DAP, an d KCI) was o u tstan ding in all three so iltr eatments , although slightly better in II than III, th e two best treatments in th elot , not significant ly di ff er ent, however. Ferti lizer B (CaCN2, CaP , and KC I) didpoorly in all three. Ferti liz er C (Ca(N 0 3 h, CaP, and KC I) was less affected by so iltreatment. Fertil izer D (ur ea, CaP , and KCI) produced rat her well in II an d III bu tpoo rly in 1. E (urea and CaP) pe rformed abou t as well in I a nd II b u t better in III.Ur ea (F ) seemed very toxic to growth . This treat m en t was repea te d t hre e differenttimes, and eac h time t he response wa s the same . Yet , when urea was used togetherwi t h calc iu m phosphate and p ot assium ch loride (D) , app aren tly t he u rea t ox icit ydis appeared even though growt h was less good th an fo r A and C. Fertilizer G(CaCN 2, CaP , an d KN0 3 ) , although no t a good co mbin ation , showed t he average

42

Soil treatment

HAWAII AGRICULTURAL EXP ERIMENT STATION

Tabl e 31. Dry weight yield s (grams pe r pot) by tr eatments

Dry weight (grams per pot)(average of 98 yield s)

IIIIII

ACDEGBF

LeachedGypsumCheck

Fertilizer tr eatments

SA, DAP, KCICa(N03h , CaP, KCIUrea, CaP, KClUrea, CaPCaCN2, CaP, KN03CaCN2, CaP, KCIUre a

HSD

HSD

5.10 (c)4.57 (b)3.85 (a)

.33

Dry weight (grams per pot)(average of 42 yields)

7.24 (e)5.63 (d)5.59 (d)4.94 (c)3.96 (b)3.35 (b)

.84 (a)

.62

Minor elem ents andmagnesium tr eatments

Dry weight (gram s per pot)(average of 42 yields)

23I7456

ZnS04CUS04CheckNa2Mo04MnS04MgS04H3B03

HSD

5.34 (c)4.58 (b)4.53 (b)4.49 (a) (b)4.35 (a) (b)4.35 (a) (b)3.89 (a)

.62

tr end with soil treatment. It is quite obvious that ap p ly ing Ca fertilizers will notsolve the problem.

Interactions between soil treatments and m in or elements we re significant.Although soil treatment III was th e best for growth, and alt ho ugh Zn ga ve thehighest yield for the minor elemen ts , the highest ac t ua l yield in Table 33 is for the

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FA ILU RE 43

Tabl e 32. Aver age dr y weight yield (grams per pot). Interacti on between soil treatments andfer t ilizers (F = 2.54 **) (top yields under scored)

Fertilizers

Soil treatment A B C D E I' G Average

I Check 6.59 2.6 8 5.26 4. 30 4. 30 .62 3. 19 3.85II Gy psum 7.63 3.35 5.73 6. 17 4. 5 6 .34 4. 19 4.57

III Leach ed 7.49 4.0 1 5. 89 6.30 5.94 1.5 6 4.49 5. 10

Averag e 7.2 4 3.25 5.63 5.59 4.9 3 .84 3.96

Table 33 . Average dry weight yield (grams per po t). Soil tr eatments X min or elem entint eractions (I' = 2.7 0** ) (top yields undersco red)

Check Zn Cu Mn Mg B Mo

Soil tr eatment 2 3 4 5 6 7 Average

I Chec k 3.88 5.35 3.55 3.63 3.36 3.08 4.10 3.85II Gypsum 4. 58 5. 8 1 4.7 6 3.92 4.7 4 3.89 4.2 8 4.57

III Leach ed 5.1 5 4.86 5.43 5.5 0 4.9 3 4.70 5. 11 5. 10

Average 4. 5 3 5.34 4. 58 4.35 4. 35 3.89 4.4 9

gy p su m-t reate d soil with Z n , but the untreat ed soi l with Zn a lso did relat ive ly we ll ,eve n better than th e lea ched so il with Zn . Leach ed so il did it s best in co m bi na tio nwith Cu an d Mn , and it s lo ss with Zn will be ex p la ined later.

T able 34 shows the fertil izers b y minor elements an d Mg interactions. The bestgro w t h was o b ta ine d in th e A -2 co m binat io n with A-7, A- I , and A-3 foll owing in

orde r. Fer tili zer s C a nd D a lso did rel ativel y well with Zn, but fertilizer A with Znwas th e best co m binat ion o f th ese two tr eatments.

Summarizing this ex perime n t, using gy p su m at th e ra te of 5 ton s per acr e ,fertilizer A a nd Zn ap peare d t o give th e best grow th b u t o n ly sligh tly better th anleaching . Leaching the so i l res u lte d in A fert ili zer doing bes t with Cu a nd Mn. No tt reating th e so il a t all , but usin g fertilize r A a nd Z n, did rel ativel y well. Such , then ,are the co m b inations from wh ich pl antation m an agement ma y select o ne or m o repracti ces for field experime n tation and co rrec t io n of MG F.

As a fin al chec k , sin ce so me of th e so il supp ly remained , several 5-q uar t p las ti cpails were fill ed , an d seve n di ff er ent fe r t ilizer co m b ina t io n s we re ap p lied to give5 00 pounds Nand 21 8 o f P per ac re . Sudangr ass aga in was u sed , and fi na l thinning

44 HAWAII AGRICULTURAL EXPERIMENT STATION

Table 34. Average dry weight yield (grams per pot). Fertilizers X minor elementsand magnesium interactions (top yields underscored)

Check Zn Cu Mn Mg B Mo

Fertilizers 2 3 4 5 6 7 Average

A SA, DAP, KCI 7.55 8.07 7.5 1 6.95 6.74 6.09 7.76 7.24B CaCN 2' CaP, KCl 3.06 3.6 1 3.56 3,41 3.19 2.98 3.62 3.35C Ca(N03b CaP, KCI 5.55 7,42 5.98 5.32 4.95 4.35 5.85 5.63D Urea, CaP, KCI 5,40 7.14 5.17 4.76 5.72 5.22 5.74 5.59E Urea, CaP 5.16 6.29 5.00 5.21 4.52 4.53 3.85 4.93F Urea .74 .8 1 .96 .92 1.04 .66 .77 .84G CaCN 2 , CaP, KN03 4.28 4.06 3.91 3.88 4.28 3.4 1 3.89 3.96

Average 4.53 5.34 4.58 4.35 4.35 3.89 4.49

Tab le 35. Average green weigh t yie ld of sudangrass (grams per po t)

FertilizersSoil

tr ea tm entsYield of sudangrass

(green weigh t per cult ure )

{ Leached 330SA, DAP, ZnS04 Gypsum 275

Chec k 318

{ Leached 351SA + DAP and all th e minor cleme nts exce pt boron Gypsum 250

Check 266

{ Leach ed 349DAP + ZnS04 Gypsum

Chec k 239

{ Leached 320SA + DAP Gypsum

Chec k 242

{ Leach ed 224SA + ZnS 0 4 Gypsum 36

Check 75

{ Leach ed 135Urea + ZnS04 Gypsum 50

Check 63

{ Leach ed 129Aqua ammo nia + ZnS04 Gy psum

Check 48

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILU RE 45

to 15 p lan t s per cu lture was effec te d. In Table 35 , the green weight yield o f thesuda ng rass as o b ta ine d is recorded alo ng with th e fe rti lize rs an d supp lements u sed.

Urea wa s included aga in, and aga in, eve n with Zn S0 4, gave ve ry poo r gro w th; infact, it produced a very to xi c co n d it io n . However, aqua am mo n ia with ZnS04 alsoproduced very poor grow t h. Using DAP fo r a ll th e N (a nd th is p rovided so me extraphosphat e ) and ZnS04 ga ve the b est gro w t h on leach ed so il alo ng with SA , DAP ,an d all the minor elements except B. SA with ZnS 04 was very inferio r to DAP withZnS04 o r DAP with SA, po inting to so me very spec if ic ro le being pl ayed e it he r bythe phosphate ion o r by the am mo n ia supp lied by a pho sphat e ion , eve n though thesoil an d pl ants wer e very ri ch in phosphate . As in the p revi ous set of d at a, u sin g SA,DAP , and ZnS04, the untreated so il gave fa irly sa t isfac to ry y ie lds .

In this study t he gy ps u m t reat me n t did not sho w up very we ll. S in ce the leach edsoi l was su pe rio r in all t rea tmen ts, t he yields fo r it are list ed a lone in Tab le 36 .

Table 36. Average gree n weight of suda ngrass (gram s per po t)on leach ed soi l with vario us fe rt ilize rs

Fe r ti lize rs

SA + DAP and all the minor cleme nts except boronDAP + ZnS0 4SA + DAP + Zn S04SA + DAP (no Zn S04)SA + ZnS04Urea + Zn S04Aqu a ammoni a + Zn S04

Yield of suda ngrass(green weig ht per cultur e)

35 1349330320224135129

Again, DAP with o r without SA but with Zn shows the best result s an dundoubtedly po ints the way toward m ax im um y ie lds in t he MG F areas, eve nthough ther e are excessive amoun ts of P already in th e pl ant and so il.

Effect of Treatments on the Absorption of the Various Nutrients by SudangrassIn o rde r to see k reasons fo r so me o f th e inter actions, a co m p le te nut rient

an aly sis o f the su da ng rass was mad e afte r th e dry weight d a ta wer e o b tain ed fo r th efir st sudang rass ser ies (Tables 3 1 to 34). The P co n te n ts by t rea t me nt are shown inTable 3 7.

Leaching the so il, a lt ho ug h giving the best grow t h, resulted in lowest P. The Zntreatment, also giving the best grow t h of th e minor e leme n ts , had the lowest P level.Most o f these differ en ces , however, ar e dilution effect s. Of a ll the fertilizer s, SA ,DAp, an d KCl gave the high es t yi eld and had the high est level of P, and t h is maywell be significant , but fertilizer B which had next to the lowest y ie ld h ad next toth e highest P. T he re seems little rel ation between the P co n te n t an d ac t ual gro w th

46 HAWAII AGRICULTURAL EXPERIMENT ST ATION

Tabl e 37 . Phosph orus co ntent of suda ngra ss b y tr ea tm ent (per cen t dry weight )

So iltreat- Perc ent Per cen t Minor Percentmcnt P Fertilizers P elem ents P

Chec k . 118 (c) A SA, DAP , KCI .14 7 (c) 4 Mn .109 (b)Gypsum .111 (b) B CaCN2, CaP , KCI .11 2 (b) 3 Cu .10 6 (b)Leached .0 79 (a) G CaCN2, CaP, KN0 3 . 108 (b) 5 Mg .10 5 (b)

HSD .005 D Urea , CaP, KCI .091 (a) 6 B .105 (b)F Urea .088 (a) 7 Mo .104 (b)E Urea , Ca P .086 (a) I Che ck .099 (a)C Ca(N03 )z, CaP, KCI .086 (a) 2 Zn .090 (a)

HSD .00 99 HSD .00 99

Ta ble 38 . Calc ium co ntent o f suda ngrass by treatm en t (percen t dry weigh t)

So iltrea t- Percent Per cen t Minor Percen tment Ca Ferti lizers Ca eleme nt Ca

Gy psum .342 (c) F Urea .373 (e) 6 B .321 (c)Leac hed .295 (b) C Ca(N0 3)z, CaP, KCI .355 (d) 5 Mg .31 3 (c)Chec k .268 (a ) E Urea, CaP .29 7 (c) I Chec k .30 7 (b) (c)

HSD .007 D Urea, CaP, KCI .283 (b) 3 Cu .301 (b) (c)B CaCN2, CaP, KCI .272 (b) 7 Mo .299 (b)A SA, DAP, KCI .270 (a) (b) 4 Mn .297 (b)G CaCN2 , CaP, KN0 3 .623 (a) 2 Zn .2 75 (b)

HSD .0 13 (a) HSD .0 13 (a)

m ade, ye t th e presence o f excess P in the m edium w as a very importan t part of theso lu t ion .

Raising t he Ca level (T able 38) b y gy ps u m an d leaching did raise the y ie ld , butthe A fert ilizer had the lowest Ca reading , and Zn, which ga ve the best y iel dincrease , had the lowest Ca level. It would see m that leaching away the excessso lub le sa lts co rrected t he Ca absorpt io n pro ble m ade q uate ly . Ure a, w hich gavest un t ed grow t h , had th e h ighest Ca .

With a p H near 8. 0 (Table 39 ) , Fe migh t we ll be in defic ient supp ly, but wi thleachin g ra is ing the ac idi ty and with fertilizer A co n tr ibu t ing to more acid ity (see

below ), prob lems shou ld no t deve lop. Urea alon e , w hich was toxic and gave verypoo r gro w t h , great ly increased the co nce n tration o f Fe in t he pl ant. T he addi t ion ofcalcium p hosp hate to ur ea with or without K resu lted in the lowest Fe level andmu ch impro ved grow t h .

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE

Tabl e 39. Iron co nte nt o f sud angrass by treatment (ppm dry matter)

Soiltr eat- Minorment ppm Fe Fertilizers ppm Fe eleme nt ppm Fe

Chec k 270 (b ) F Urea 49 5 (c) 6 B 266 (b)Gy psum 26 1 (b) B CaCN2, CaP, KCI 25 1 (b) 5 Mg 260 (b)Leached ill (a) G CaCN2, CaP, KN03 199 (a) (b) 1 Chec k 240 (a) (b)

HSD 30 A SA, DAP, KCl 192 (a) 3 Cu 23 7 (a) (b)C Ca(N03 l2 , CaP , KCI 180 (a) 7 Mo 22 1 (a) (b)E Urea, CaP 166 (a) 2 Zn 210 (a) (b)D Urea, CaP, KCI ill (a) 4 Mn 204 (a)

HSD 56 HSD 56

Table 40. Sulfur co nten t of sudangrass by tre atment (percent dr y matter)

Soi ltreat- Per cent Percent Minor Per centment S Ferti lizers S clem ent S

Gy psum .1 22 (b) B CaCN2 , CaP, KCl .085 (b) 5 Mg .088 (b)Leached .05 6 (a) G CaCN2, CaP, KN0 3 .085 (b) 6 B .0 79 (a) (b)Check .054 (a) A SA, DAP, KCl .083 (b) 3 Cu .0 78 (a) (b)

HSD .00 5 C Ca(N03 l2, CaP, KCI .075 (a) (b) 4 Mn .07 6 (a)D Urea, CaP, KCI .0 73 (a) 2 Zn .073 (a)F Urea .0 72 (a) 1 Chec k .073 (a)E Urea, CaP .06 8 (a) 7 Mo .0 72 (a)

HSD .0 10 HSD .010

47

Le ach ing had no effec t on S (Table 40) , but gypsum did, and if S defi cien cydo es occur as a part o f the gro wt h failure co m plex , th en leaching by its elf wouldfail to correct th e problem en tirely . The use of SA raised th e S significantly, eventh ough rather little. Mg, B, an d Cu ea ch raised the S levels.

The only o u tstandi ng effec t on B (Table 41) was that of the addit ion of B whichmore than tripl ed the level s in the plants , which probably accounts for thesignifican t loss o f yield (Table 3 1). Lea ching lowered the B levels significantly butrel atively little.

Reducing the level o f K (Tab le 4 2) is desir able be cau se o f the high amou n ts o fth e elem ent in th e soil. Leach ing, of cou rse , lower ed it the mo st, but gy ps u m alsolow ered it, though not so effec t ivel y . Fertilizer s E, D, A , an d C also lowered it.Minor elements had no effec t.

48 HAWAII AG RICULT URA L EXPE RIMENT STATION

Ta b le 4 1. Boron co nte nt of suda ngra ss b y treatmen t (ppm dry matt er)

Soi ltrea t- ppm ppm Minor ppmment B Fer t ilizers 13 cle ment s B

Check 27. 17 F Urea 29 .2 (b) (c) 6 13 74. 1 (b)Gy psum 26. 16 (a) (b) A SA, OAP, KCI 27 .6 (b) (c) 7 Mo 19. 1 (a)Leach ed 24 .3 1 (a) 13 CaCN 2, CaP, KCI 26.7 (a) (b) (c) 3 Cu 18.5 (a )

HSO 2.2 1 G CaCN 2, Ca P, KN0 3 25.3 (a) (b ) (c) 5 Mg 17.7 (a)E Urea , CaP 24.9 (a) (b ) 4 Mn 17.7 (a)0 Urea, CaP, KCl 24.5 (a) (b) 1 Chec k 17.0 (a)C Ca(N03 h , CaP, KCI 22.9 (a) 2 Zn .!L.Q (a)

HSO 4.2 (a) HSO 4. 2

Tabl e 42. Potassium co nte nt o f sudangrass by tre atment (per cent dry matter)

Soi ltr ea t- Per cen t Percent Mino r Percent

ment K Fer tilizers K elements K

Check 2.84 (c) F Urea 3.30 (c) 6 13 2.64Gypsum 2.68 (b) G CaCN2, CaP, KN0 3 2.68 (b) 7 Mo 2.63Leach ed 2.26 (a) B CaCN2 , CaP, KCl 2.65 (b) 3 Cu 2.61

HSO .08 E Urea , CaP 2.44 (a) 4 Mn 2.600 Urea, CaP, KCI 2.39 (a) I Check 2.60A SA, OAP , KCI 2.38 (a) 5 Mg 2.58C Ca(N0 3h, Ca P, KCI 2.33 (a) 2 Zn 2.5 0

HSO . 16 n.s .

S ince a Zn ap p lica t ion improved growth in th e chec k and gy ps um tr eatments butno t in the leach ed soil (Table 4 3) , it is inter esting that , so me how, leach ing ena b ledth e p lant to abs orb mu ch more Zn . It might be sugg est ed that, just as Na , perhap stogether with K, pr evented the ready abso rp tio n of Ca, so they also depressed theabsorp tion of Zn. Becau se gy ps u m fai led to increase Zn uptak e very much, andbecau se th e A fert ilizer a lso fail ed to affect Zn uptak e , th e ap p lica t ion o f Zn toth ese tr eatments re sulted in a marked st imu la t ion o f growt h .

Soil pH Ch angesTo determine the effect on so il pH w hich various co mmerc ial fer t ilize rs could

have, quantiti es of so il from Haw aii an Commercial an d S uga r Co mpany were p lacedin to 5-q ua rt pl astic pail s without a p rovi sion for dr ainage. Amo u nts of fertilizer

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE

Table 43. Zinc conte nt o f suda ngrass by trea tmen t (ppm dry matter)

Soilt reat- pp m ppm Minor ppmment Zn Fe rtilizers Zn eleme nts Zn

Leached 44.68 (c) F Urea 57 .0 (c) 2 Zn 40.1 (b)Gypsum 28.98 (b) E Urea, CaP 32.8 (b ) 6 B 33 .8 (a)Chec k 27 .52 (a) D Urea, CaP, KCl 31.7 (b) 7 Mo 33 .1 (a)

HSD 1.25 C Ca(N0312, CaP, KCI 30.5 (a) (b ) 3 Cu 32.9 (a)B CaCN2 , CaP, KCI 28.8 (a) (b) I Check 32 .8 (a)G CaCN2, CaP, KN0 3 28.0 (a) 5 Mg 32.0 (a)A SA, DAP , KCI 27.5 (a) 4 Mn 31.6 (a)

HSD 3.7 HSD 3.7

49

adde d provided N and P in ver y large amoun ts , equiva lent to abo u t 12 crop s. SoilpH readings were tak en at various intervals of time, but are re porte d for on ly three(Ta b le 44). Afte r abo u t 8 months and wi th th e untreat ed so il showing a pH of 7.4 2 ,th e most ac id fer tilizer was ammoniu m nitrat e and DAP , followed in order b y SA,DAP , and KCl ; SA and treble superphosphate; 16-16-16; tripolyphos ; ammoniumnitrate and am m op hos; ammonium chlor ide and DAP; DAP alo ne ; and am mo niu m

Table 44. Effec t of various fertilizers on pH of MGF soil (ori ginal pH - 7.50 )

pH at:

Ferti lizers

CheckCaCN2Non ammonium fer tilize rKN03 and treble superphosph at e and Ca(N0 3hAmm onium chloride and tr eble superphospha teDiammon ium ph osph at eSA and ammo phosAmmo nium chlori de and DAPAmmo nium nitrat e and tr ebl e superphos pha teTri polyphos16-16-16SA and tr ebl e supe rphosphateSA, DAP, and KCIAmmo nium nitrat e and DAP

8/30/68

7.508.006.7 86.405.606.805.5 96.105. 275 .455.555.2 46.1 45.8 0

12/24/68

7.357.806.306.2 05.655.3 14.5 35.204.954. 874.104.9 24.7 04.73

. 4/ 17/6 9

7.4 27.1 26.856.1 55.124.424.404.354.304. 254. 194. 144.083.72

50 HAWAII AGRICULTURAL EXPERIMENT STATION

chloride and treble superpho sphate. Calcium nitrate , potassium nitrate, and tr eblesup erp hos phate ; a nonammonium mixture (made up of Ca and K nitrates with Kphosphates and KCl); and ca lciu m cyanam id had some ac id ify ing qualities, a lthoug hth ey would normall y be expecte d to add basicity.

Undoubtedly, one reason fertilizer A (SA , DAP , and KCl) es sen tia lly ov er cameMaui gro wt h failure was not o n ly the addit ion of th e elem ents themselves but alsothe cha nge in pll which would have a marked effec t on nutri ent absorption .

General Conclusions and RecommendationsIt see ms rather cle ar that Maui growth fa ilure is th e result o f several interacting

factors , an d th at severa l ave n ues are ava ilable for it s co rrect ion. High level s of Na aswell as K have tended to red uce the Ca as we ll as the Zn and su lfa te levels of thepl ant s. Leaching was supe rio r to gy ps u m in correct ing t he monovalent ca tio nproblem, but gy psu m increased both th e Ca and sulfa te up tak e perhap s more thannecessary . The use of am mo n iu m su lfa te and di ammonium pho sphate , or th e latteralone, not only tended to aci d if y the so il but also pl ayed so me unknown role in theabso rp t io n process, whi ch imp ro ved th e Ca uptak e. Becau se lea ching somehowgreatly increased th e level of Zn in the pl ants, this e leme n t was important only onthe untreat ed and gy ps u m-treate d soi l. Leaching, howev er , resulted in need fo raddit ions o f Mn and perhaps a lso Cu. B was definitel y harmful , perhap s indicatingso me th ing of an excess in th e na tural soil.

Choices by man ageme nt fo r th e co rrec tio n of MG F are severa l. Leaching wouldundoubtedly have the most perman ent ef fec t, particularl y if the pump wat ernormally used co u ld be im proved in quality. It wo u ld not m atter wh ether t his isdone during winter months whe n mountain wa ter supp lies a re excessive or duringreg ular crop production by using fresh mountain water for irri gation . Eve nirr iga t ing during the winter months with mountain water would b e helpful.Actually , the fields involved are irr igated en tire ly with water pumped from withinth e MG F a rea it self, resulting in a re cycling o f the affec te d water. Perhaps, also ,

some st ud ies might be undertak en to use excess mounta in water in th e winter todilute th e ground wat er sup p lies b y direct flow downward into th e gro und w at erbasin, thus usin g the underground as a massive fr esh water reservoir for mountainwater.

G ypsum may also find a place , parti cul arl y in lo cali zed areas sh own to be low intissu e Ca and S. The use o f di ammonium pho sphat e , perhaps a superi o r N so urc e ,see ms so o uts tanding in th e expe rime n ts reported that, even though phosphatelevel s in ca ne alr eady see m high, an intensive program of fie ld demonstration an dstudy shou ld be undert ak en. Even using no soil treatment at a ll, but usingdiammonium phosphat e for all th e N and adding Zn, would seem ce rt ain to behelpful. If lea ching is done, perhaps Zn would not be needed, but th en atten tionshould be fo cused on the needs for Mn an d Cu.

CAUSES OF LEAF FRECKLE, LEAF CHLOROSIS, AND GROWTH FAILURE

REFERENCES CITED

51

(1) CLEMENTS, HARRY F. 1951. Crop-logging sU!r"r cane in wet areas. Hawaiian SugarTechno!. 1951 Rep. P. 10-17.

(2) . 1961. The coral stone-phosphate experiment s on the Hilo and HamakuaCoasts. Hawaiian Sugar Tcchnol. 1961 Rep. P. 77-105.

(3) . 1965 . Effects of silicate on the growth and leaf freckle of sugarcane in Hawaii.Proc. 12th Congr.lnt. Soc. Sugar Cane Te chno!. P. 197-215.

(4) . 1965 . The roles of ca lcium silicate slags in sugar cane growth. Hawaiian SugarTechnol, 1965 Rep. P. 103-126.

(5) . Unpublished photographs.

(6) , E. W. pm'MAN, and j. R. WILSON . 1967. Eliminating soil toxicities withcalcium metasilicate. Hawaiian Sugar Technol. 1967 Rep. P. 43-54.

(7) , G. SHIGEURA, and E. K. AKAMINE. 1952. Factors affecting the growth ofsugar cane . Hawaii Agr, Exp. Sta. Te ch. Bull. 18.90 p.

(8) FOX, R . L.,]. A. SILVA, D. L. PLUCKNETT, and D. Y. TERANISHl. 1969. So luble andtotal silicon in sugar cane. Plant and Soil 30:81-92.

(9) HALAIS, P. 1966. Le silicate de chaux, un novel amendement pour la regeneration dessols ferrallitiques seniles en vue de culture de la canne. Mauritius Sugar Industry Res. Inst,Circ. 19. 12 p.

(10) and D. H. PARISH. 1963 . Silic a and manganese con tents of cane leaf sheaths inrelation to soil and nutrition. Mauritius Sugar Industry Res . Inst. Annu, Rep. 1963. P.74-76.

(I I) KA NEHIRO, Y., and M. F. HADANO. 1959. Effect of nitrogen ferti lizers on soil pH .Hawaii Farm Sci . 8(2) :4-5.

(12) MACDONALD, GORDON A. 1968. Composition and origin of Hawaiian lavas. Geolog.Soc. Arner. Memoir 116. P. 477-522.

(13) MAHILUM, B. C., R. L. FOX, and] . A. SILVA. 1970. Residual effects of liming volcanicash soils in the humid tropics. Soil Sci. 109 :102-109.

(14) MARTIN,]. P. 1938. Sugar cane diseases in Hawaii. Exp. Sta., Hawaiian Sugar Planters'Assoc., Honolulu, Hawaii. xiv + 295 p.

(15) MATSUSAKA, YOSHITO, and G. D. SHERMAN. 1950. Titration curves and bufferingcapacities of Hawaiian soils. Hawaii Agr. Exp. Sta, Tech. Bull. 11. 36 p.

(16) MCGEORGE, W. T. 1923. The acidity of Hawaiian soil. Hawaiian Planters' Rec,27:215-226.

(17) . 1924. Iron, aluminum and manganese in the soil solution of Hawaiian soils.Hawaiian Planters' Rec. 28:294-303.

(18) 1924. Soil acidity: its relation to root-rot. Hawaiian Planters' Rec . 28:321-324.

(19) . 1925 . The influence of aluminum, manganese and iron salts upon the growthof sugar cane, and their relation to the infer t ility of acid island soil. Exp. Sta.; HawaiianSugar Plan ters' Assoc. Agr . and Chern. Ser. Bull. 49. 95 p.

52 HAWAII AGRICULTURAL EXPERIMENT STATION

(20) MONTEITH, N. H. , and G . D. SHERMAN. 1963. The compa ra t ive effec ts of ca lciumcarbonate and of ca lcium silica te on the yield of Sudan grass grown in a FerruginousLatosol and a Hydrol Humic Latoso l. Hawaii Agr . Exp. St a. T ech . Bull. 53. 40 p.

(2 1) MOTOMURA, S. 1962. Effect of organic matters on th e formation of ferrous iro n in so ils.Soil Science and Plant Nu trition (.Jap an) 8:20- 29 .

(2 2) OADES, J. M. 1963. The nature and distribution of iron co mpou nds in soils. So ils andFertilizers 26: 69-80 .

(2 3) OKAWA, K. 1936. Investigation on th e ph ysiological ac t ion of silici c acid for plants.J.Sci. Soil and Manure (.Japan) 10:95-110,216-243,414-428.

(24) PIERRE, W. If. 1928. Nitrogen ous fertilizers and soi l acidity: 1. Effect of var iousnitrogenous fertilizers on soil rea ction . .J . Am er, So c. Agron . 20:254-269 .

(25) . 19 33. Det ermination of equivalent ac id ity and bas icit y of fertilizers. Indust .and Engin. Ch ern. (Analytical Ed.) 5:229-234 .

(26) PONNAMPERUMA, F. N. 1964. Dynamic asp ects of flooded soi ls and th e nutrition o f th eri ce plant. In The mineral nutrition o f th e rice plant. Pro c. of a sy mposium at th eInternation al Rice Research In stitute, Los Ban os, Philippines, 1964. Th e Johns HopkinsPress, Baltimore. P. 295-328.

(27) RIXON, A. .J . 1966. Effects o f heavy applica t ion of lime to soils derived from volcanic ashon the humid Hilo and Hamakua Coa st s, Island of Hawaii. Hawaii Agr. Exp. Sta. Tech.Bull. 47. 52 p .

(28) SAMUELS, G . , and A. G. ALEXANDER. 1968. Influen ce of variable manganese andsilicon on th e nutrition, sugar production and enz yme act ivity of immature sugarcane.Pro c, 13th Congr., Int. Soc. Sugar Can e Techno!' P. 544-555.

(29) and F. GONZALEZ-VELEZ. 1962. The influence of ammonium sulpha tefertilization on the pH of sugarcane soils. J. Agr. Univ, Puerto Rico. 46:297-306.

(30) SUEHISA, R. H. , O. R. YOUNGE, and G. D. SHERMAN. 1963. Effect s of silicat es on th eph osphorus availabilit y to Sudan grass gro wn in Hawaii an soils. Hawaii Agr. Exp . St a .Te ch. Bull. 51. 40 p.

(31) WANG, THOMAS S. C., SAN -YAO CHENG , and HELEN TUNG. 1967 . Extraction andanalysi s of so il o rgan ic ac ids . Soil Sci. 103: 360- 366.

(3 2) _138-144.

_ ___, and . 1967. Dynamics o f so il orga nic ac ids . Soi l Sci . 104:

(33) and TZE-TANG CHUANG. 1967. Soil alco ho ls, the ir dynamics and th eir effectupon plant gro wt h. So il Sci. 104:40-45 .

(34) , TZE·KEN YANG, and SANoYAO CHENG. 1967. Am ino acids in subtropicalsoi l hydrolysates. Soi l Sci. 103:67 -74.

(35) WILLIAMS, D. E., and J. VLAMI S. 1957. The ef fec t of silicon on yield and man gan ese-54uptake and di stribution in th e leaves o f barl ey plant s grown in cu ltu re solu t ions . PlantPhysiol. 32: 40 4-409 .

Single cop ies o f this publication availab le without ch arge to Hawaii residents fromcou n ty agen ts. Out-of-State inquiries or bulk orde rs should be sent to the College ofTropical Agr iculture Order De sk, Room 108 Krau ss Hall, 25 0 0 Dol e Street,Honolulu, Haw aii 9682 2. Pri ce per copy to bulk users, forty-five cen ts plus postage.

Hawaii Agricultural Experiment StationCollege of Tropical Agriculture , University of HawaiiC. Peairs Wilson , Dean of the College and Director of the Experiment StationLeslie D. Swindale, Associate Director of the Experiment Station

Tech. Bull. 88-February 1974 (3.5M)